Skip to contentSkip to conversation

5.17.15-rc1 review

PATCH In review Greg Kroah-Hartman 6 participants 312 messages Jun 13, 2022 – Aug 23, 2023 Page 1 of 3

Subsystems (4)
lore
Share & tools
lore

Greg Kroah-Hartman

author maintainer
#0 lore raw

cover 5.17.15-rc1 review

This is the start of the stable review cycle for the 5.17.15 release. There are 298 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 Wed, 15 Jun 2022 09:47:08 +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-5.17.15-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-5.17.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 5.17.15-rc1

Damien Le Moal <[email protected]>
zonefs: fix handling of explicit_open option on mount

Johan Hovold <[email protected]>
PCI: qcom: Fix pipe clock imbalance

Christoph Hellwig <[email protected]>
block, loop: support partitions without scanning

Davide Caratti <[email protected]>
net/sched: act_police: more accurate MTU policing

Pascal Hambourg <[email protected]>
md/raid0: Ignore RAID0 layout if the second zone has only one device

Jason A. Donenfeld <[email protected]>
random: account for arch randomness in bits

Jason A. Donenfeld <[email protected]>
random: mark bootloader randomness code as __init

Jason A. Donenfeld <[email protected]>
random: avoid checking crng_ready() twice in random_init()

KuoHsiang Chou <[email protected]>
drm/ast: Create threshold values for AST2600

Michael Ellerman <[email protected]>
powerpc/32: Fix overread/overwrite of thread_struct via ptrace

Jason Wang <[email protected]>
virtio-rng: make device ready before making request

Alex Deucher <[email protected]>
drm/amdgpu: update VCN codec support for Yellow Carp

Mohammad Zafar Ziya <[email protected]>
drm/amdgpu/jpeg2: Add jpeg vmid update under IB submit

Brian Norris <[email protected]>
drm/atomic: Force bridge self-refresh-exit on CRTC switch

Brian Norris <[email protected]>
drm/bridge: analogix_dp: Support PSR-exit to disable transition

Michael Ellerman <[email protected]>
powerpc: Don't select HAVE_IRQ_EXIT_ON_IRQ_STACK

Matthew Wilcox (Oracle) <[email protected]>
mm/huge_memory: Fix xarray node memory leak

Peter Zijlstra <[email protected]>
cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE

Xie Yongji <[email protected]>
vduse: Fix NULL pointer dereference on sysfs access

Mathias Nyman <[email protected]>
Input: bcm5974 - set missing URB_NO_TRANSFER_DMA_MAP urb flag

Olivier Matz <[email protected]>
ixgbe: fix unexpected VLAN Rx in promisc mode on VF

Olivier Matz <[email protected]>
ixgbe: fix bcast packets Rx on VF after promisc removal

Martin Faltesek <[email protected]>
nfc: st21nfca: fix incorrect sizing calculations in EVT_TRANSACTION

Martin Faltesek <[email protected]>
nfc: st21nfca: fix memory leaks in EVT_TRANSACTION handling

Martin Faltesek <[email protected]>
nfc: st21nfca: fix incorrect validating logic in EVT_TRANSACTION

Jchao Sun <[email protected]>

    writeback: Fix inode->i_io_list not be protected by inode->i_lock error

Ilya Maximets <[email protected]>
net: openvswitch: fix misuse of the cached connection on tuple changes

Tan Tee Min <[email protected]>
net: phy: dp83867: retrigger SGMII AN when link change

Adrian Hunter <[email protected]>
mmc: block: Fix CQE recovery reset success

Sergey Shtylyov <[email protected]>
ata: libata-transport: fix {dma|pio|xfer}_mode sysfs files

Tyler Erickson <[email protected]>
libata: fix translation of concurrent positioning ranges

Tyler Erickson <[email protected]>
libata: fix reading concurrent positioning ranges log

David Safford <[email protected]>
KEYS: trusted: tpm2: Fix migratable logic

Tyler Erickson <[email protected]>
scsi: sd: Fix interpretation of VPD B9h length

Shyam Prasad N <[email protected]>
cifs: populate empty hostnames for extra channels

Paulo Alcantara <[email protected]>
cifs: fix reconnect on smb3 mount types

Shyam Prasad N <[email protected]>
cifs: return errors during session setup during reconnects

Jeremy Soller <[email protected]>
ALSA: hda/realtek: Add quirk for HP Dev One

Cameron Berkenpas <[email protected]>
ALSA: hda/realtek: Fix for quirk to enable speaker output on the Lenovo Yoga DuetITL 2021

huangwenhui <[email protected]>
ALSA: hda/conexant - Fix loopback issue with CX20632

Takashi Iwai <[email protected]>
ALSA: usb-audio: Set up (implicit) sync for Saffire 6

Takashi Iwai <[email protected]>
ALSA: usb-audio: Skip generic sync EP parse for secondary EP

Bedant Patnaik <[email protected]>
platform/x86: hp-wmi: Use zero insize parameter only when supported

Jorge Lopez <[email protected]>

    platform/x86: hp-wmi: Fix hp_wmi_read_int() reporting error (0x05)

Kuan-Ying Lee <[email protected]>
scripts/gdb: change kernel config dumping method

Jiasheng Jiang <[email protected]>
platform/x86: barco-p50-gpio: Add check for platform_driver_register

Xie Yongji <[email protected]>
vringh: Fix loop descriptors check in the indirect cases

Kees Cook <[email protected]>
nodemask: Fix return values to be unsigned

Yury Norov <[email protected]>
drm/amd/pm: use bitmap_{from,to}_arr32 where appropriate

Steve French <[email protected]>
cifs: version operations for smb20 unneeded when legacy support disabled

Christian Borntraeger <[email protected]>
s390/gmap: voluntarily schedule during key setting

Bjorn Helgaas <[email protected]>
Revert "PCI: brcmstb: Split brcm_pcie_setup() into two funcs"

Bjorn Helgaas <[email protected]>
Revert "PCI: brcmstb: Add mechanism to turn on subdev regulators"

Bjorn Helgaas <[email protected]>
Revert "PCI: brcmstb: Add control of subdevice voltage regulators"

Bjorn Helgaas <[email protected]>
Revert "PCI: brcmstb: Do not turn off WOL regulators on suspend"

Yu Kuai <[email protected]>
nbd: fix io hung while disconnecting device

Yu Kuai <[email protected]>
nbd: fix race between nbd_alloc_config() and module removal

Yu Kuai <[email protected]>
nbd: call genl_unregister_family() first in nbd_cleanup()

Peter Zijlstra <[email protected]>
jump_label,noinstr: Avoid instrumentation for JUMP_LABEL=n builds

Peter Zijlstra <[email protected]>
x86/cpu: Elide KCSAN for cpu_has() and friends

Masahiro Yamada <[email protected]>
modpost: fix undefined behavior of is_arm_mapping_symbol()

Johannes Berg <[email protected]>
um: line: Use separate IRQs per line

Lijo Lazar <[email protected]>
drm/amd/pm: Fix missing thermal throttler status

Gong Yuanjun <[email protected]>
drm/amd/pm: fix a potential gpu_metrics_table memory leak

Gong Yuanjun <[email protected]>
drm/radeon: fix a possible null pointer dereference

David Galiffi <[email protected]>
drm/amd/display: Check if modulo is 0 before dividing.

Daniel Borkmann <[email protected]>
net, neigh: Set lower cap for neigh_managed_work rearming

Xiubo Li <[email protected]>
ceph: flush the mdlog for filesystem sync

Venky Shankar <[email protected]>
ceph: allow ceph.dir.rctime xattr to be updatable

Michal Kubecek <[email protected]>
Revert "net: af_key: add check for pfkey_broadcast in function pfkey_process"

Oder Chiou <[email protected]>
ASoC: rt5640: Do not manipulate pin "Platform Clock" if the "Platform Clock" is not in the DAPM

Hannes Reinecke <[email protected]>
scsi: myrb: Fix up null pointer access on myrb_cleanup()

Guoqing Jiang <[email protected]>
md: protect md_unregister_thread from reentrancy

Hyunchul Lee <[email protected]>
ksmbd: smbd: fix connection dropped issue

Liu Xinpeng <[email protected]>
watchdog: wdat_wdt: Stop watchdog when rebooting the system

Hao Luo <[email protected]>
kernfs: Separate kernfs_pr_cont_buf and rename_lock.

John Ogness <[email protected]>
serial: msm_serial: disable interrupts in __msm_console_write()

Wang Cheng <[email protected]>

    staging: rtl8712: fix uninit-value in r871xu_drv_init()

Wang Cheng <[email protected]>
staging: rtl8712: fix uninit-value in usb_read8() and friends

Andre Przywara <[email protected]>
clocksource/drivers/sp804: Avoid error on multiple instances

bumwoo lee <[email protected]>
extcon: Modify extcon device to be created after driver data is set

Dan Carpenter <[email protected]>
extcon: Fix extcon_get_extcon_dev() error handling

Shuah Khan <[email protected]>
misc: rtsx: set NULL intfdata when probe fails

Srinivas Kandagatla <[email protected]>
soundwire: qcom: adjust autoenumeration timeout

Heikki Krogerus <[email protected]>
usb: dwc3: host: Stop setting the ACPI companion

Marek Szyprowski <[email protected]>

    usb: dwc2: gadget: don't reset gadget's driver->bus

Changbin Du <[email protected]>
sysrq: do not omit current cpu when showing backtrace of all active CPUs

Hangyu Hua <[email protected]>
char: xillybus: fix a refcount leak in cleanup_dev()

Evan Green <[email protected]>
USB: hcd-pci: Fully suspend across freeze/thaw cycle

Duoming Zhou <[email protected]>
drivers: usb: host: Fix deadlock in oxu_bus_suspend()

Duoming Zhou <[email protected]>
drivers: tty: serial: Fix deadlock in sa1100_set_termios()

Zhen Ni <[email protected]>
USB: host: isp116x: check return value after calling platform_get_resource()

Duoming Zhou <[email protected]>
drivers: staging: rtl8192e: Fix deadlock in rtllib_beacons_stop()

Duoming Zhou <[email protected]>
drivers: staging: rtl8192u: Fix deadlock in ieee80211_beacons_stop()

Mika Westerberg <[email protected]>
thunderbolt: Use different lane for second DisplayPort tunnel

Huang Guobin <[email protected]>
tty: Fix a possible resource leak in icom_probe

Zheyu Ma <[email protected]>
tty: synclink_gt: Fix null-pointer-dereference in slgt_clean()

Duoming Zhou <[email protected]>
drivers: staging: rtl8192bs: Fix deadlock in rtw_joinbss_event_prehandle()

Duoming Zhou <[email protected]>
drivers: staging: rtl8723bs: Fix deadlock in rtw_surveydone_event_callback()

Kees Cook <[email protected]>
lkdtm/usercopy: Expand size of "out of frame" object

Miquel Raynal <[email protected]>
iio: st_sensors: Add a local lock for protecting odr

Xiaoke Wang <[email protected]>
staging: rtl8712: fix a potential memory leak in r871xu_drv_init()

Xiaoke Wang <[email protected]>
iio: dummy: iio_simple_dummy: check the return value of kstrdup()

David Howells <[email protected]>
iov_iter: Fix iter_xarray_get_pages{,_alloc}()

Etienne van der Linde <[email protected]>
nfp: flower: restructure flow-key for gre+vlan combination

Linus Torvalds <[email protected]>
drm: imx: fix compiler warning with gcc-12

Muchun Song <[email protected]>
tcp: use alloc_large_system_hash() to allocate table_perturb

Marek Behún <[email protected]>
net: dsa: mv88e6xxx: use BMSR_ANEGCOMPLETE bit for filling an_complete

Miaoqian Lin <[email protected]>
net: altera: Fix refcount leak in altera_tse_mdio_create

Willem de Bruijn <[email protected]>
ip_gre: test csum_start instead of transport header

Mark Bloch <[email protected]>
net/mlx5: fs, fail conflicting actions

Feras Daoud <[email protected]>
net/mlx5: Rearm the FW tracer after each tracer event

Saeed Mahameed <[email protected]>
net/mlx5: Fix mlx5_get_next_dev() peer device matching

Mark Bloch <[email protected]>
net/mlx5: Lag, filter non compatible devices

Masahiro Yamada <[email protected]>
net: ipv6: unexport __init-annotated seg6_hmac_init()

Masahiro Yamada <[email protected]>
net: xfrm: unexport __init-annotated xfrm4_protocol_init()

Masahiro Yamada <[email protected]>
net: mdio: unexport __init-annotated mdio_bus_init()

Chuck Lever <[email protected]>

    SUNRPC: Fix the calculation of xdr->end in xdr_get_next_encode_buffer()

Maciej Fijalkowski <[email protected]>
xsk: Fix handling of invalid descriptors in XSK TX batching API

Magnus Karlsson <[email protected]>
i40e: xsk: Move tmp desc array from driver to pool

Gal Pressman <[email protected]>
net/mlx4_en: Fix wrong return value on ioctl EEPROM query failure

Miaoqian Lin <[email protected]>
net: dsa: lantiq_gswip: Fix refcount leak in gswip_gphy_fw_list

Eric Dumazet <[email protected]>

    bpf, arm64: Clear prog->jited_len along prog->jited

Jan Beulich <[email protected]>
x86: drop bogus "cc" clobber from __try_cmpxchg_user_asm()

Kuniyuki Iwashima <[email protected]>
af_unix: Fix a data-race in unix_dgram_peer_wake_me().

Christophe JAILLET <[email protected]>
stmmac: intel: Fix an error handling path in intel_eth_pci_probe()

Masahiro Yamada <[email protected]>
xen: unexport __init-annotated xen_xlate_map_ballooned_pages()

Miaoqian Lin <[email protected]>
net: ethernet: bgmac: Fix refcount leak in bcma_mdio_mii_register

Taehee Yoo <[email protected]>
amt: fix wrong type string definition

Taehee Yoo <[email protected]>
amt: fix possible null-ptr-deref in amt_rcv()

Taehee Yoo <[email protected]>
amt: fix wrong usage of pskb_may_pull()

Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: bail out early if hardware offload is not supported

Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: memleak flow rule from commit path

Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: release new hooks on unsupported flowtable flags

Miaoqian Lin <[email protected]>
ata: pata_octeon_cf: Fix refcount leak in octeon_cf_probe

Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: always initialize flowtable hook list in transaction

Chuck Lever <[email protected]>
SUNRPC: Trap RDMA segment overflows

Michael Ellerman <[email protected]>
powerpc/kasan: Force thread size increase with KASAN

Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: delete flowtable hooks via transaction list

Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: use kfree_rcu(ptr, rcu) to release hooks in clean_net path

Florian Westphal <[email protected]>
netfilter: nat: really support inet nat without l3 address

Steven Price <[email protected]>
drm/panfrost: Job should reference MMU not file_priv

Marek Vasut <[email protected]>

    drm/bridge: ti-sn65dsi83: Handle dsi_lanes == 0 as invalid

Kinglong Mee <[email protected]>
xprtrdma: treat all calls not a bcall when bc_serv is NULL

Chao Yu <[email protected]>
f2fs: fix to tag gcing flag on page during file defragment

Daniel Bristot de Oliveira <[email protected]>
rtla/Makefile: Properly handle dependencies

Greg Ungerer <[email protected]>
m68knommu: fix undefined reference to `mach_get_rtc_pll'

Liao Chang <[email protected]>
RISC-V: use memcpy for kexec_file mode

Yang Yingliang <[email protected]>
video: fbdev: pxa3xx-gcu: release the resources correctly in pxa3xx_gcu_probe/remove()

Saurabh Sengar <[email protected]>
video: fbdev: hyperv_fb: Allow resolutions with size > 64 MB for Gen1

Trond Myklebust <[email protected]>
NFSv4: Don't hold the layoutget locks across multiple RPC calls

Radhey Shyam Pandey <[email protected]>
dmaengine: zynqmp_dma: In struct zynqmp_dma_chan fix desc_size data type

Greg Ungerer <[email protected]>
m68knommu: fix undefined reference to `_init_sp'

Greg Ungerer <[email protected]>
m68knommu: set ZERO_PAGE() to the allocated zeroed page

Lucas Tanure <[email protected]>
i2c: cadence: Increase timeout per message if necessary

Dongliang Mu <[email protected]>
f2fs: remove WARN_ON in f2fs_is_valid_blkaddr

Yang Yingliang <[email protected]>
iommu/arm-smmu-v3: check return value after calling platform_get_resource()

Yang Yingliang <[email protected]>
iommu/arm-smmu: fix possible null-ptr-deref in arm_smmu_device_probe()

Mark-PK Tsai <[email protected]>
tracing: Avoid adding tracer option before update_tracer_options

Jun Miao <[email protected]>
tracing: Fix sleeping function called from invalid context on RT kernel

Jeff Xie <[email protected]>
tracing: Make tp_printk work on syscall tracepoints

Masami Hiramatsu <[email protected]>
bootconfig: Make the bootconfig.o as a normal object file

Gong Yuanjun <[email protected]>
mips: cpc: Fix refcount leak in mips_cpc_default_phys_base

Dave Jiang <[email protected]>
dmaengine: idxd: set DMA_INTERRUPT cap bit

Linus Torvalds <[email protected]>
bluetooth: don't use bitmaps for random flag accesses

Luiz Augusto von Dentz <[email protected]>
Bluetooth: hci_sync: Fix attempting to suspend with unfiltered passive scan

Luiz Augusto von Dentz <[email protected]>
Bluetooth: MGMT: Add conditions for setting HCI_CONN_FLAG_REMOTE_WAKEUP

Leo Yan <[email protected]>
perf c2c: Fix sorting in percent_rmt_hitm_cmp()

Zhengjun Xing <[email protected]>
perf record: Support sample-read topdown metric group for hybrid platforms

Kan Liang <[email protected]>
perf parse-events: Move slots event for the hybrid platform too

Kan Liang <[email protected]>
perf evsel: Fixes topdown events in a weak group for the hybrid platform

Saravana Kannan <[email protected]>
driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Hoang Le <[email protected]>
tipc: check attribute length for bearer name

Duoming Zhou <[email protected]>
ax25: Fix ax25 session cleanup problems

Damien Le Moal <[email protected]>
scsi: sd: Fix potential NULL pointer dereference

Kuogee Hsieh <[email protected]>
drm/msm/dp: Always clear mask bits to disable interrupts at dp_ctrl_reset_irq_ctrl()

David Howells <[email protected]>
afs: Fix infinite loop found by xfstest generic/676

Haibo Chen <[email protected]>
gpio: pca953x: use the correct register address to do regcache sync

Fabien Parent <[email protected]>
regulator: mt6315-regulator: fix invalid allowed mode

Alexander Gordeev <[email protected]>
s390/mcck: isolate SIE instruction when setting CIF_MCCK_GUEST flag

Ziyang Xuan <[email protected]>
macsec: fix UAF bug for real_dev

Dan Carpenter <[email protected]>
octeontx2-af: fix error code in is_valid_offset()

Jason Wang <[email protected]>
vdpa: ifcvf: set pci driver data in probe

Eric Dumazet <[email protected]>
tcp: tcp_rtx_synack() can be called from process context

Guoju Fang <[email protected]>
net: sched: add barrier to fix packet stuck problem for lockless qdisc

Maxim Mikityanskiy <[email protected]>
net/mlx5e: Update netdev features after changing XDP state

Changcheng Liu <[email protected]>
net/mlx5: correct ECE offset in query qp output

Maxim Mikityanskiy <[email protected]>
net/mlx5e: Disable softirq in mlx5e_activate_rq to avoid race condition

Paul Blakey <[email protected]>
net/mlx5: CT: Fix header-rewrite re-use for tupels

Maor Dickman <[email protected]>
net/mlx5e: TC NIC mode, fix tc chains miss table

Leon Romanovsky <[email protected]>
net/mlx5: Don't use already freed action pointer

Christophe JAILLET <[email protected]>
virtio: pci: Fix an error handling path in vp_modern_probe()

Eli Cohen <[email protected]>
vdpa: Fix error logic in vdpa_nl_cmd_dev_get_doit

Haisu Wang <[email protected]>
blk-mq: do not update io_ticks with passthrough requests

Jens Axboe <[email protected]>
block: make bioset_exit() fully resilient against being called twice

Íñigo Huguet <[email protected]>
sfc: fix wrong tx channel offset with efx_separate_tx_channels

Martin Habets <[email protected]>
sfc: fix considering that all channels have TX queues

Christoph Hellwig <[email protected]>
block: use bio_queue_enter instead of blk_queue_enter in bio_poll

Yu Xiao <[email protected]>
nfp: only report pause frame configuration for physical device

Eric Dumazet <[email protected]>

    tcp: add accessors to read/set tp->snd_cwnd

Guangguan Wang <[email protected]>
net/smc: fixes for converting from "struct smc_cdc_tx_pend **" to "struct smc_wr_tx_pend_priv *"

Heinrich Schuchardt <[email protected]>
riscv: read-only pages should not be writable

Yu Kuai <[email protected]>
nbd: don't clear 'NBD_CMD_INFLIGHT' flag if request is not completed

Christoph Hellwig <[email protected]>
block: take destination bvec offsets into account in bio_copy_data_iter

Menglong Dong <[email protected]>
bpf: Fix probe read error in ___bpf_prog_run()

Song Liu <[email protected]>
selftests/bpf: fix stacktrace_build_id with missing kprobe/urandom_read

Andrii Nakryiko <[email protected]>
selftests/bpf: fix selftest after random: Urandom_read tracepoint removal

Zhihao Cheng <[email protected]>
ubi: ubi_create_volume: Fix use-after-free when volume creation failed

Zhihao Cheng <[email protected]>
ubi: fastmap: Fix high cpu usage of ubi_bgt by making sure wl_pool not empty

Baokun Li <[email protected]>
jffs2: fix memory leak in jffs2_do_fill_super

Genjian Zhang <[email protected]>
ep93xx: clock: Do not return the address of the freed memory

Alexander Lobakin <[email protected]>
modpost: fix removing numeric suffixes

Miaoqian Lin <[email protected]>
net: dsa: mv88e6xxx: Fix refcount leak in mv88e6xxx_mdios_register

Miaoqian Lin <[email protected]>
net: ethernet: ti: am65-cpsw-nuss: Fix some refcount leaks

Dan Carpenter <[email protected]>
net: ethernet: mtk_eth_soc: out of bounds read in mtk_hwlro_get_fdir_entry()

Vincent Ray <[email protected]>
net: sched: fixed barrier to prevent skbuff sticking in qdisc backlog

Michael Walle <[email protected]>
net: lan966x: check devm_of_phy_get() for -EDEFER_PROBE

Dan Carpenter <[email protected]>
drm/amdgpu: Off by one in dm_dmub_outbox1_low_irq()

Eddie James <[email protected]>
spi: fsi: Fix spurious timeout

liuyacan <[email protected]>

    net/smc: set ini->smcrv2.ib_dev_v2 to NULL if SMC-Rv2 is unavailable

Taehee Yoo <[email protected]>
amt: fix possible memory leak in amt_rcv()

Taehee Yoo <[email protected]>
amt: fix return value of amt_update_handler()

Jann Horn <[email protected]>
s390/crypto: fix scatterwalk_unmap() callers in AES-GCM

Krzysztof Kozlowski <[email protected]>
clocksource/drivers/oxnas-rps: Fix irq_of_parse_and_map() return value

Shengjiu Wang <[email protected]>
ASoC: fsl_sai: Fix FSL_SAI_xDR/xFR definition

Ming Lei <[email protected]>

    blk-mq: don't touch ->tagset in blk_mq_get_sq_hctx

Miaoqian Lin <[email protected]>
watchdog: ts4800_wdt: Fix refcount leak in ts4800_wdt_probe

Miaoqian Lin <[email protected]>
watchdog: rti-wdt: Fix pm_runtime_get_sync() error checking

Zhang Wensheng <[email protected]>
driver core: fix deadlock in __device_attach

Schspa Shi <[email protected]>
driver: base: fix UAF when driver_attach failed

Tony Lindgren <[email protected]>
bus: ti-sysc: Fix warnings for unbind for serial

Miaoqian Lin <[email protected]>
firmware: dmi-sysfs: Fix memory leak in dmi_sysfs_register_handle

Ilpo Järvinen <[email protected]>
serial: stm32-usart: Correct CSIZE, bits, and parity

Ilpo Järvinen <[email protected]>
serial: st-asc: Sanitize CSIZE and correct PARENB for CS7

Ilpo Järvinen <[email protected]>
serial: sifive: Sanitize CSIZE and c_iflag

Ilpo Järvinen <[email protected]>
serial: sh-sci: Don't allow CS5-6

Ilpo Järvinen <[email protected]>
serial: txx9: Don't allow CS5-6

Ilpo Järvinen <[email protected]>
serial: rda-uart: Don't allow CS5-6

Ilpo Järvinen <[email protected]>
serial: digicolor-usart: Don't allow CS5-6

Ilpo Järvinen <[email protected]>
serial: uartlite: Fix BRKINT clearing

YueHaibing <[email protected]>
serial: cpm_uart: Fix build error without CONFIG_SERIAL_CPM_CONSOLE

Ilpo Järvinen <[email protected]>

    serial: 8250_fintek: Check SER_RS485_RTS_* only with RS485

John Ogness <[email protected]>

    serial: meson: acquire port->lock in startup()

Jiasheng Jiang <[email protected]>
staging: r8188eu: add check for kzalloc

Miaoqian Lin <[email protected]>
rtc: ftrtc010: Fix error handling in ftrtc010_rtc_probe

Yang Yingliang <[email protected]>
rtc: mt6397: check return value after calling platform_get_resource()

Howard Chiu <[email protected]>
ARM: dts: aspeed: ast2600-evb: Enable RX delay for MAC0/MAC1

Samuel Holland <[email protected]>
clocksource/drivers/riscv: Events are stopped during CPU suspend

Miaoqian Lin <[email protected]>
soc: rockchip: Fix refcount leak in rockchip_grf_init

Nícolas F. R. A. Prado <[email protected]>
dt-bindings: remoteproc: mediatek: Make l1tcm reg exclusive to mt819x

Li Jun <[email protected]>
extcon: ptn5150: Add queue work sync before driver release

Xin Xiong <[email protected]>
ksmbd: fix reference count leak in smb_check_perm_dacl()

Guilherme G. Piccoli <[email protected]>
coresight: cpu-debug: Replace mutex with mutex_trylock on panic notifier

Pierre-Louis Bossart <[email protected]>
soundwire: intel: prevent pm_runtime resume prior to system suspend

Biju Das <[email protected]>
watchdog: rzg2l_wdt: Fix reset control imbalance

Biju Das <[email protected]>
watchdog: rzg2l_wdt: Fix 'BUG: Invalid wait context'

Biju Das <[email protected]>
watchdog: rzg2l_wdt: Fix Runtime PM usage

Biju Das <[email protected]>
watchdog: rzg2l_wdt: Fix 32bit overflow issue

Greg Kroah-Hartman <[email protected]>
export: fix string handling of namespace in EXPORT_SYMBOL_NS

Maciej W. Rozycki <[email protected]>
serial: sifive: Report actual baud base rather than fixed 115200

Linus Walleij <[email protected]>
power: supply: ab8500_fg: Allocate wq in probe

Hans de Goede <[email protected]>
power: supply: axp288_fuel_gauge: Drop BIOS version check from "T3 MRD" DMI quirk

Johan Hovold <[email protected]>
phy: qcom-qmp: fix pipe-clock imbalance on power-on failure

Guilherme G. Piccoli <[email protected]>
misc/pvpanic: Convert regular spinlock into trylock on panic path

Krzysztof Kozlowski <[email protected]>
rpmsg: qcom_smd: Fix returning 0 if irq_of_parse_and_map() fails

Cixi Geng <[email protected]>
iio: adc: sc27xx: Fine tune the scale calibration values

Cixi Geng <[email protected]>
iio: adc: sc27xx: fix read big scale voltage not right

Miaoqian Lin <[email protected]>

    iio: proximity: vl53l0x: Fix return value check of wait_for_completion_timeout

Miaoqian Lin <[email protected]>
iio: adc: stmpe-adc: Fix wait_for_completion_timeout return value check

Arnaud Pouliquen <[email protected]>
rpmsg: virtio: Fix the unregistration of the device rpmsg_ctrl

Hangyu Hua <[email protected]>
rpmsg: virtio: Fix possible double free in rpmsg_virtio_add_ctrl_dev()

Hangyu Hua <[email protected]>
rpmsg: virtio: Fix possible double free in rpmsg_probe()

Bjorn Andersson <[email protected]>
usb: typec: mux: Check dev_set_name() return value

Xiaomeng Tong <[email protected]>
firmware: stratix10-svc: fix a missing check on list iterator

Xiaomeng Tong <[email protected]>
misc: fastrpc: fix an incorrect NULL check on list iterator

SeongJae Park <[email protected]>
scripts/get_abi: Fix wrong script file name in the help message

Zheng Yongjun <[email protected]>
usb: dwc3: pci: Fix pm_runtime_get_sync() error checking

Wesley Cheng <[email protected]>
usb: dwc3: gadget: Replace list_for_each_entry_safe() if using giveback

Krzysztof Kozlowski <[email protected]>
rpmsg: qcom_smd: Fix irq_of_parse_and_map() return value

Uwe Kleine-König <[email protected]>
pwm: raspberrypi-poe: Fix endianness in firmware struct

Uwe Kleine-König <[email protected]>
pwm: lp3943: Fix duty calculation in case period was clamped

Christophe JAILLET <[email protected]>
staging: fieldbus: Fix the error handling path in anybuss_host_common_probe()

Miaoqian Lin <[email protected]>
usb: musb: Fix missing of_node_put() in omap2430_probe

Lin Ma <[email protected]>
USB: storage: karma: fix rio_karma_init return

Niels Dossche <[email protected]>
usb: usbip: add missing device lock on tweak configuration cmd

Hangyu Hua <[email protected]>
usb: usbip: fix a refcount leak in stub_probe()

Samuel Holland <[email protected]>
phy: rockchip-inno-usb2: Fix muxed interrupt support

Peng Fan <[email protected]>
remoteproc: imx_rproc: Ignore create mem entry for resource table

Sherry Sun <[email protected]>
tty: serial: fsl_lpuart: fix potential bug when using both of_alias_get_id and ida_simple_get

Miaoqian Lin <[email protected]>
serial: 8250_aspeed_vuart: Fix potential NULL dereference in aspeed_vuart_probe

Daniel Gibson <[email protected]>
tty: n_tty: Restore EOF push handling behavior

Miaoqian Lin <[email protected]>
tty: serial: owl: Fix missing clk_disable_unprepare() in owl_uart_probe

Wang Weiyang <[email protected]>
tty: goldfish: Use tty_port_destroy() to destroy port

Christophe Leroy <[email protected]>
lkdtm/bugs: Don't expect thread termination without CONFIG_UBSAN_TRAP

Jiasheng Jiang <[email protected]>
lkdtm/bugs: Check for the NULL pointer after calling kmalloc

Alexandru Tachici <[email protected]>
iio: adc: ad7124: Remove shift from scan_type

Jakob Koschel <[email protected]>
staging: greybus: codecs: fix type confusion of list iterator variable

Randy Dunlap <[email protected]>
pcmcia: db1xxx_ss: restrict to MIPS_DB1XXX boards

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

Diffstat:

 Documentation/ABI/testing/sysfs-ata                |  11 +-
 .../bindings/regulator/mt6315-regulator.yaml       |   4 +-
 .../devicetree/bindings/remoteproc/mtk,scp.yaml    |  44 ++--
 Documentation/tools/rtla/Makefile                  |  14 +-
 Makefile                                           |   4 +-
 arch/arm/boot/dts/aspeed-ast2600-evb.dts           |   4 +-
 arch/arm/mach-ep93xx/clock.c                       |  10 +-
 arch/arm64/net/bpf_jit_comp.c                      |   1 +
 arch/m68k/Kconfig.machine                          |   1 +
 arch/m68k/include/asm/pgtable_no.h                 |   3 +-
 arch/m68k/kernel/setup_mm.c                        |   7 -
 arch/m68k/kernel/setup_no.c                        |   1 -
 arch/m68k/kernel/time.c                            |   9 +
 arch/mips/kernel/mips-cpc.c                        |   1 +
 arch/powerpc/Kconfig                               |   2 -
 arch/powerpc/include/asm/thread_info.h             |  10 +-
 arch/powerpc/kernel/ptrace/ptrace-fpu.c            |  20 +-
 arch/powerpc/kernel/ptrace/ptrace.c                |   3 +
 arch/riscv/kernel/efi.c                            |   2 +-
 arch/riscv/kernel/machine_kexec.c                  |   4 +-
 arch/s390/crypto/aes_s390.c                        |   4 +-
 arch/s390/kernel/entry.S                           |   6 +-
 arch/s390/mm/gmap.c                                |  14 ++
 arch/um/drivers/chan_kern.c                        |  10 +-
 arch/um/drivers/line.c                             |  22 +-
 arch/um/drivers/line.h                             |   4 +-
 arch/um/drivers/ssl.c                              |   2 -
 arch/um/drivers/stdio_console.c                    |   2 -
 arch/um/include/asm/irq.h                          |  22 +-
 arch/x86/include/asm/cpufeature.h                  |   2 +-
 arch/x86/include/asm/uaccess.h                     |   2 +-
 block/bio.c                                        |   9 +-
 block/blk-core.c                                   |   2 +-
 block/blk-mq.c                                     |  10 +-
 block/genhd.c                                      |   2 +
 drivers/ata/libata-core.c                          |  21 +-
 drivers/ata/libata-scsi.c                          |   2 +-
 drivers/ata/libata-transport.c                     |   2 +-
 drivers/ata/pata_octeon_cf.c                       |   3 +
 drivers/base/bus.c                                 |   4 +-
 drivers/base/dd.c                                  |  10 +-
 drivers/block/loop.c                               |   8 +-
 drivers/block/nbd.c                                |  55 +++--
 drivers/bus/ti-sysc.c                              |   4 +-
 drivers/char/hw_random/virtio-rng.c                |   2 +
 drivers/char/random.c                              |  15 +-
 drivers/char/xillybus/xillyusb.c                   |   1 +
 drivers/clocksource/timer-oxnas-rps.c              |   2 +-
 drivers/clocksource/timer-riscv.c                  |   2 +-
 drivers/clocksource/timer-sp804.c                  |  10 +-
 drivers/dma/idxd/dma.c                             |   1 +
 drivers/dma/xilinx/zynqmp_dma.c                    |   5 +-
 drivers/extcon/extcon-axp288.c                     |   4 +-
 drivers/extcon/extcon-ptn5150.c                    |  11 +
 drivers/extcon/extcon.c                            |  33 +--
 drivers/firmware/dmi-sysfs.c                       |   2 +-
 drivers/firmware/stratix10-svc.c                   |  12 +-
 drivers/gpio/gpio-pca953x.c                        |  19 +-
 drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c             |   6 +-
 drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.h             |   1 +
 drivers/gpu/drm/amd/amdgpu/nv.c                    |   1 +
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c  |   2 +-
 .../gpu/drm/amd/display/dc/dce/dce_clock_source.c  |   9 +-
 drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c     |   2 +-
 drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c |   1 +
 drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c     |   2 +-
 .../gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c   |   3 +
 drivers/gpu/drm/ast/ast_mode.c                     |   5 +-
 drivers/gpu/drm/bridge/analogix/analogix_dp_core.c |  42 +++-
 drivers/gpu/drm/bridge/ti-sn65dsi83.c              |   2 +-
 drivers/gpu/drm/drm_atomic_helper.c                |  16 +-
 drivers/gpu/drm/imx/ipuv3-crtc.c                   |   2 +-
 drivers/gpu/drm/msm/dp/dp_ctrl.c                   |   9 +-
 drivers/gpu/drm/panfrost/panfrost_drv.c            |   5 +-
 drivers/gpu/drm/panfrost/panfrost_job.c            |   6 +-
 drivers/gpu/drm/panfrost/panfrost_job.h            |   2 +-
 drivers/gpu/drm/radeon/radeon_connectors.c         |   4 +
 drivers/hwtracing/coresight/coresight-cpu-debug.c  |   7 +-
 drivers/i2c/busses/i2c-cadence.c                   |  12 +-
 drivers/idle/intel_idle.c                          |  32 ++-
 drivers/iio/adc/ad7124.c                           |   1 -
 drivers/iio/adc/sc27xx_adc.c                       |  20 +-
 drivers/iio/adc/stmpe-adc.c                        |   8 +-
 drivers/iio/common/st_sensors/st_sensors_core.c    |  24 +-
 drivers/iio/dummy/iio_simple_dummy.c               |  20 +-
 drivers/iio/proximity/vl53l0x-i2c.c                |   7 +-
 drivers/input/mouse/bcm5974.c                      |   7 +-
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c        |   2 +
 drivers/iommu/arm/arm-smmu/arm-smmu.c              |   5 +-
 drivers/md/md.c                                    |  15 +-
 drivers/md/raid0.c                                 |  31 +--
 drivers/misc/cardreader/rtsx_usb.c                 |   1 +
 drivers/misc/fastrpc.c                             |   9 +-
 drivers/misc/lkdtm/bugs.c                          |  10 +-
 drivers/misc/lkdtm/lkdtm.h                         |   8 +-
 drivers/misc/lkdtm/usercopy.c                      |  17 +-
 drivers/misc/pvpanic/pvpanic.c                     |  10 +-
 drivers/mmc/core/block.c                           |   3 +-
 drivers/mtd/ubi/fastmap-wl.c                       |  69 ++++--
 drivers/mtd/ubi/fastmap.c                          |  11 -
 drivers/mtd/ubi/ubi.h                              |   4 +-
 drivers/mtd/ubi/vmt.c                              |   1 -
 drivers/mtd/ubi/wl.c                               |  19 +-
 drivers/net/amt.c                                  |  59 +++--
 drivers/net/dsa/lantiq_gswip.c                     |   4 +-
 drivers/net/dsa/mv88e6xxx/chip.c                   |   1 +
 drivers/net/dsa/mv88e6xxx/serdes.c                 |  27 +--
 drivers/net/ethernet/altera/altera_tse_main.c      |   6 +-
 drivers/net/ethernet/broadcom/bgmac-bcma-mdio.c    |   1 +
 drivers/net/ethernet/intel/i40e/i40e_txrx.c        |  11 -
 drivers/net/ethernet/intel/i40e/i40e_txrx.h        |   1 -
 drivers/net/ethernet/intel/i40e/i40e_xsk.c         |   4 +-
 drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c     |   8 +-
 .../net/ethernet/marvell/octeontx2/af/rvu_cpt.c    |   2 +-
 drivers/net/ethernet/mediatek/mtk_eth_soc.c        |   3 +
 drivers/net/ethernet/mellanox/mlx4/en_ethtool.c    |   2 +-
 drivers/net/ethernet/mellanox/mlx5/core/dev.c      |  72 ++++--
 .../ethernet/mellanox/mlx5/core/diag/fw_tracer.c   |   7 +-
 drivers/net/ethernet/mellanox/mlx5/core/en.h       |   4 +
 drivers/net/ethernet/mellanox/mlx5/core/en/fs.h    |   2 +
 drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c   |   1 +
 .../ethernet/mellanox/mlx5/core/en/reporter_rx.c   |   6 +
 drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c |  19 +-
 drivers/net/ethernet/mellanox/mlx5/core/en/trap.c  |   1 +
 .../net/ethernet/mellanox/mlx5/core/en/xsk/pool.c  |   1 +
 .../net/ethernet/mellanox/mlx5/core/en/xsk/setup.c |   5 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c  |  29 ++-
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c    |  38 ++-
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c  |  37 ++-
 drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c  |  12 +-
 .../net/ethernet/mellanox/mlx5/core/mlx5_core.h    |   1 +
 .../ethernet/mellanox/mlx5/core/steering/fs_dr.c   |   9 +-
 .../net/ethernet/microchip/lan966x/lan966x_main.c  |   9 +-
 .../net/ethernet/netronome/nfp/flower/conntrack.c  |  32 +--
 drivers/net/ethernet/netronome/nfp/flower/match.c  |  16 +-
 .../net/ethernet/netronome/nfp/nfp_net_ethtool.c   |   4 +-
 drivers/net/ethernet/sfc/efx_channels.c            |   6 +-
 drivers/net/ethernet/sfc/net_driver.h              |   2 +-
 drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c  |   4 +-
 drivers/net/ethernet/ti/am65-cpsw-nuss.c           |   3 +-
 drivers/net/macsec.c                               |   7 +
 drivers/net/phy/dp83867.c                          |  29 +++
 drivers/net/phy/mdio_bus.c                         |   1 -
 drivers/nfc/st21nfca/se.c                          |  53 +++--
 drivers/pci/controller/dwc/pcie-qcom.c             |   6 -
 drivers/pci/controller/pcie-brcmstb.c              | 257 +++------------------
 drivers/pcmcia/Kconfig                             |   2 +-
 drivers/phy/qualcomm/phy-qcom-qmp.c                |   2 +-
 drivers/phy/rockchip/phy-rockchip-inno-usb2.c      |  10 +-
 drivers/platform/x86/barco-p50-gpio.c              |   5 +-
 drivers/platform/x86/hp-wmi.c                      |  23 +-
 drivers/power/supply/ab8500_fg.c                   |  19 +-
 drivers/power/supply/axp288_charger.c              |  17 +-
 drivers/power/supply/axp288_fuel_gauge.c           |   1 -
 drivers/power/supply/charger-manager.c             |   7 +-
 drivers/power/supply/max8997_charger.c             |   8 +-
 drivers/pwm/pwm-lp3943.c                           |   1 +
 drivers/pwm/pwm-raspberrypi-poe.c                  |   2 +-
 drivers/remoteproc/imx_rproc.c                     |   3 +
 drivers/rpmsg/qcom_smd.c                           |   4 +-
 drivers/rpmsg/virtio_rpmsg_bus.c                   |   9 +-
 drivers/rtc/rtc-ftrtc010.c                         |  34 ++-
 drivers/rtc/rtc-mt6397.c                           |   2 +
 drivers/scsi/myrb.c                                |  11 +-
 drivers/scsi/sd.c                                  |   3 +-
 drivers/soc/rockchip/grf.c                         |   2 +
 drivers/soundwire/intel.c                          |   3 +
 drivers/soundwire/qcom.c                           |   2 +-
 drivers/spi/spi-fsi.c                              |  12 +-
 drivers/staging/fieldbus/anybuss/host.c            |   2 +-
 drivers/staging/greybus/audio_codec.c              |   4 +-
 drivers/staging/r8188eu/core/rtw_xmit.c            |  13 +-
 drivers/staging/r8188eu/include/rtw_xmit.h         |   2 +-
 drivers/staging/rtl8192e/rtllib_softmac.c          |   2 +-
 .../staging/rtl8192u/ieee80211/ieee80211_softmac.c |   2 +-
 drivers/staging/rtl8712/os_intfs.c                 |   1 -
 drivers/staging/rtl8712/usb_intf.c                 |  12 +-
 drivers/staging/rtl8712/usb_ops.c                  |  27 ++-
 drivers/staging/rtl8723bs/core/rtw_mlme.c          |  12 +-
 drivers/thunderbolt/tb.c                           |  19 +-
 drivers/thunderbolt/test.c                         |  16 +-
 drivers/thunderbolt/tunnel.c                       |  11 +-
 drivers/thunderbolt/tunnel.h                       |   4 +-
 drivers/tty/goldfish.c                             |   2 +
 drivers/tty/n_tty.c                                |  38 ++-
 drivers/tty/serial/8250/8250_aspeed_vuart.c        |   2 +
 drivers/tty/serial/8250/8250_fintek.c              |   8 +-
 drivers/tty/serial/cpm_uart/cpm_uart_core.c        |   2 +-
 drivers/tty/serial/digicolor-usart.c               |   2 +
 drivers/tty/serial/fsl_lpuart.c                    |  24 +-
 drivers/tty/serial/icom.c                          |   2 +-
 drivers/tty/serial/meson_uart.c                    |  13 ++
 drivers/tty/serial/msm_serial.c                    |   5 +
 drivers/tty/serial/owl-uart.c                      |   1 +
 drivers/tty/serial/rda-uart.c                      |   2 +
 drivers/tty/serial/sa1100.c                        |   4 +-
 drivers/tty/serial/serial_txx9.c                   |   2 +
 drivers/tty/serial/sh-sci.c                        |   6 +-
 drivers/tty/serial/sifive.c                        |   8 +-
 drivers/tty/serial/st-asc.c                        |   4 +
 drivers/tty/serial/stm32-usart.c                   |  15 +-
 drivers/tty/serial/uartlite.c                      |   3 +-
 drivers/tty/synclink_gt.c                          |   2 +
 drivers/tty/sysrq.c                                |  13 +-
 drivers/usb/core/hcd-pci.c                         |   4 +-
 drivers/usb/dwc2/gadget.c                          |   1 -
 drivers/usb/dwc3/drd.c                             |   9 +-
 drivers/usb/dwc3/dwc3-pci.c                        |   2 +-
 drivers/usb/dwc3/gadget.c                          |  20 +-
 drivers/usb/dwc3/host.c                            |   2 -
 drivers/usb/host/isp116x-hcd.c                     |   6 +-
 drivers/usb/host/oxu210hp-hcd.c                    |   2 +
 drivers/usb/musb/omap2430.c                        |   1 +
 drivers/usb/phy/phy-omap-otg.c                     |   4 +-
 drivers/usb/storage/karma.c                        |  15 +-
 drivers/usb/typec/mux.c                            |  14 +-
 drivers/usb/typec/tcpm/fusb302.c                   |   4 +-
 drivers/usb/usbip/stub_dev.c                       |   2 +-
 drivers/usb/usbip/stub_rx.c                        |   2 +
 drivers/vdpa/ifcvf/ifcvf_main.c                    |   3 +-
 drivers/vdpa/vdpa.c                                |  13 +-
 drivers/vdpa/vdpa_user/vduse_dev.c                 |   7 +-
 drivers/vhost/vringh.c                             |  10 +-
 drivers/video/fbdev/hyperv_fb.c                    |  19 +-
 drivers/video/fbdev/pxa3xx-gcu.c                   |  12 +-
 drivers/virtio/virtio_pci_modern_dev.c             |   1 +
 drivers/watchdog/rti_wdt.c                         |   2 +-
 drivers/watchdog/rzg2l_wdt.c                       |  46 ++--
 drivers/watchdog/ts4800_wdt.c                      |   5 +-
 drivers/watchdog/wdat_wdt.c                        |   1 +
 drivers/xen/xlate_mmu.c                            |   1 -
 fs/afs/dir.c                                       |   5 +-
 fs/ceph/mds_client.c                               |  33 ++-
 fs/ceph/xattr.c                                    |  10 +-
 fs/cifs/cifsfs.c                                   |   2 +-
 fs/cifs/cifsfs.h                                   |   2 +-
 fs/cifs/cifsglob.h                                 |   4 +-
 fs/cifs/connect.c                                  |   4 +
 fs/cifs/misc.c                                     |  27 ++-
 fs/cifs/sess.c                                     |   5 +-
 fs/cifs/smb2ops.c                                  |   7 +-
 fs/cifs/smb2pdu.c                                  |   3 +
 fs/f2fs/checkpoint.c                               |   4 +-
 fs/f2fs/file.c                                     |   1 +
 fs/fs-writeback.c                                  |  37 ++-
 fs/inode.c                                         |   2 +-
 fs/jffs2/fs.c                                      |   1 +
 fs/kernfs/dir.c                                    |  31 ++-
 fs/ksmbd/smbacl.c                                  |   1 +
 fs/ksmbd/transport_rdma.c                          |   1 +
 fs/nfs/nfs4proc.c                                  |   4 +
 fs/zonefs/super.c                                  |  11 +-
 include/linux/export.h                             |   7 +-
 include/linux/extcon.h                             |   2 +-
 include/linux/genhd.h                              |   1 +
 include/linux/iio/common/st_sensors.h              |   3 +
 include/linux/jump_label.h                         |   4 +-
 include/linux/mlx5/mlx5_ifc.h                      |   5 +-
 include/linux/nodemask.h                           |  38 +--
 include/linux/random.h                             |   2 +-
 include/linux/xarray.h                             |   1 +
 include/net/ax25.h                                 |   1 +
 include/net/bluetooth/hci_core.h                   |  17 +-
 include/net/flow_offload.h                         |   1 +
 include/net/netfilter/nf_tables.h                  |   1 -
 include/net/netfilter/nf_tables_offload.h          |   2 +-
 include/net/sch_generic.h                          |  42 ++--
 include/net/tcp.h                                  |  19 +-
 include/net/xdp_sock_drv.h                         |   5 +-
 include/net/xsk_buff_pool.h                        |   1 +
 include/trace/events/tcp.h                         |   2 +-
 kernel/bpf/core.c                                  |  14 +-
 kernel/trace/trace.c                               |  13 +-
 kernel/trace/trace_syscalls.c                      |  35 +--
 lib/Makefile                                       |   2 +-
 lib/iov_iter.c                                     |  20 +-
 lib/nodemask.c                                     |   4 +-
 lib/xarray.c                                       |   5 +-
 mm/huge_memory.c                                   |   3 +-
 net/ax25/af_ax25.c                                 |  27 ++-
 net/ax25/ax25_dev.c                                |   1 +
 net/ax25/ax25_subr.c                               |   2 +-
 net/bluetooth/hci_core.c                           |   4 +-
 net/bluetooth/hci_request.c                        |   2 +-
 net/bluetooth/hci_sync.c                           |  62 +++--
 net/bluetooth/mgmt.c                               |  45 ++--
 net/core/filter.c                                  |   2 +-
 net/core/flow_offload.c                            |   6 +
 net/core/neighbour.c                               |   2 +-
 net/ipv4/inet_hashtables.c                         |  10 +-
 net/ipv4/ip_gre.c                                  |  11 +-
 net/ipv4/tcp.c                                     |   8 +-
 net/ipv4/tcp_bbr.c                                 |  20 +-
 net/ipv4/tcp_bic.c                                 |  14 +-
 net/ipv4/tcp_cdg.c                                 |  30 +--
 net/ipv4/tcp_cong.c                                |  18 +-
 net/ipv4/tcp_cubic.c                               |  22 +-
 net/ipv4/tcp_dctcp.c                               |  11 +-
 net/ipv4/tcp_highspeed.c                           |  18 +-
 net/ipv4/tcp_htcp.c                                |  10 +-
 net/ipv4/tcp_hybla.c                               |  18 +-
 net/ipv4/tcp_illinois.c                            |  12 +-
 net/ipv4/tcp_input.c                               |  36 +--
 net/ipv4/tcp_ipv4.c                                |   2 +-
 net/ipv4/tcp_lp.c                                  |   6 +-
 net/ipv4/tcp_metrics.c                             |  12 +-
 net/ipv4/tcp_nv.c                                  |  24 +-
 net/ipv4/tcp_output.c                              |  34 +--
 net/ipv4/tcp_rate.c                                |   2 +-
 net/ipv4/tcp_scalable.c                            |   4 +-
 net/ipv4/tcp_vegas.c                               |  21 +-
 net/ipv4/tcp_veno.c                                |  24 +-
 net/ipv4/tcp_westwood.c                            |   3 +-
 net/ipv4/tcp_yeah.c                                |  30 +--
 net/ipv4/xfrm4_protocol.c                          |   1 -
 net/ipv6/seg6_hmac.c                               |   1 -
 net/ipv6/tcp_ipv6.c                                |   2 +-
 net/key/af_key.c                                   |  10 +-
 net/netfilter/nf_tables_api.c                      |  54 ++---
 net/netfilter/nf_tables_offload.c                  |  23 +-
 net/netfilter/nft_nat.c                            |   3 +-
 net/openvswitch/actions.c                          |   6 +
 net/openvswitch/conntrack.c                        |   4 +-
 net/sched/act_police.c                             |  16 +-
 net/smc/af_smc.c                                   |   1 +
 net/smc/smc_cdc.c                                  |   2 +-
 net/sunrpc/xdr.c                                   |   6 +-
 net/sunrpc/xprtrdma/rpc_rdma.c                     |   5 +
 net/sunrpc/xprtrdma/svc_rdma_rw.c                  |   4 +-
 net/tipc/bearer.c                                  |   3 +-
 net/unix/af_unix.c                                 |   2 +-
 net/xdp/xsk.c                                      |  16 +-
 net/xdp/xsk_buff_pool.c                            |   7 +
 net/xdp/xsk_queue.h                                |  14 +-
 scripts/gdb/linux/config.py                        |   6 +-
 scripts/get_abi.pl                                 |   4 +-
 scripts/mod/modpost.c                              |   5 +-
 security/keys/trusted-keys/trusted_tpm2.c          |   4 +-
 sound/pci/hda/patch_conexant.c                     |   7 +
 sound/pci/hda/patch_realtek.c                      |   2 +
 sound/soc/codecs/rt5640.c                          |  11 +-
 sound/soc/codecs/rt5640.h                          |   2 +
 sound/soc/fsl/fsl_sai.h                            |   4 +-
 sound/soc/intel/boards/bytcr_rt5640.c              |   2 +
 sound/usb/pcm.c                                    |   5 +-
 sound/usb/quirks-table.h                           |   7 +-
 tools/perf/arch/x86/util/evlist.c                  |   5 +-
 tools/perf/arch/x86/util/evsel.c                   |  24 +-
 tools/perf/arch/x86/util/evsel.h                   |   7 +
 tools/perf/arch/x86/util/topdown.c                 |  46 ++--
 tools/perf/arch/x86/util/topdown.h                 |   7 +
 tools/perf/builtin-c2c.c                           |   4 +-
 .../selftests/bpf/progs/test_stacktrace_build_id.c |  12 +-
 .../testing/selftests/net/forwarding/tc_police.sh  |  52 +++++
 tools/testing/selftests/netfilter/nft_nat.sh       |  43 ++++
 tools/tracing/rtla/Makefile                        |  35 +++
 356 files changed, 2368 insertions(+), 1550 deletions(-)
author maintainer
#1 lore raw

PATCH 1/298 pcmcia: db1xxx_ss: restrict to MIPS_DB1XXX boards

From: Randy Dunlap <[email protected]>

[ Upstream commit 3928cf08334ed895a31458cbebd8d4ec6d84c080 ]

When the MIPS_ALCHEMY board selection is MIPS_XXS1500 instead of MIPS_DB1XXX, the PCMCIA driver 'db1xxx_ss' has build errors due to missing DB1XXX symbols. The PCMCIA driver should be restricted to MIPS_DB1XXX instead of MIPS_ALCHEMY to fix this build error.

ERROR: modpost: "bcsr_read" [drivers/pcmcia/db1xxx_ss.ko] undefined!
ERROR: modpost: "bcsr_mod" [drivers/pcmcia/db1xxx_ss.ko] undefined!

Fixes 42a4f17dc356 ("MIPS: Alchemy: remove SOC_AU1X00 in favor of MIPS_ALCHEMY")
Cc Arnd Bergmann <[email protected]>
Cc Daniel Vetter <[email protected]>
Cc Kees Cook <[email protected]>
Cc Thomas Bogendoerfer <[email protected]>
---
 drivers/pcmcia/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pcmcia/Kconfig +1 -1
index ab53eab635f6..1740a63b814d 100644
@@ -151,7 +151,7 @@ config TCIC
 
 config PCMCIA_ALCHEMY_DEVBOARD
 	tristate "Alchemy Db/Pb1xxx PCMCIA socket services"
-	depends on MIPS_ALCHEMY && PCMCIA
+	depends on MIPS_DB1XXX && PCMCIA
 	help
 	  Enable this driver of you want PCMCIA support on your Alchemy
 	  Db1000, Db/Pb1100, Db/Pb1500, Db/Pb1550, Db/Pb1200, DB1300
Signature · 1 line
2.35.1
author maintainer
#2 lore raw

PATCH 2/298 staging: greybus: codecs: fix type confusion of list iterator variable

In reply to Greg Kroah-Hartman

From: Jakob Koschel <[email protected]>

[ Upstream commit 84ef256550196bc06e6849a34224c998b45bd557 ]

If the list does not exit early then data == NULL and 'module' does not point to a valid list element.
Using 'module' in such a case is not valid and was therefore removed.

---
 drivers/staging/greybus/audio_codec.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/staging/greybus/audio_codec.c +2 -2
index b589cf6b1d03..e19b91e7a72e 100644
@@ -599,8 +599,8 @@ static int gbcodec_mute_stream(struct snd_soc_dai *dai, int mute, int stream)
 			break;
 	}
 	if (!data) {
-		dev_err(dai->dev, "%s:%s DATA connection missing\n",
-			dai->name, module->name);
+		dev_err(dai->dev, "%s DATA connection missing\n",
+			dai->name);
 		mutex_unlock(&codec->lock);
 		return -ENODEV;
 	}
Signature · 1 line
2.35.1
author maintainer
#3 lore raw

PATCH 3/298 iio: adc: ad7124: Remove shift from scan_type

In reply to Greg Kroah-Hartman

From: Alexandru Tachici <[email protected]>

[ Upstream commit fe78ccf79b0e29fd6d8dc2e2c3b0dbeda4ce3ad8 ]

The 24 bits data is stored in 32 bits in BE. There is no need to shift it. This confuses user-space apps.

---
 drivers/iio/adc/ad7124.c | 1 -
 1 file changed, 1 deletion(-)
drivers/iio/adc/ad7124.c +0 -1
index b400bbe291aa..c22d8dcaa100 100644
@@ -188,7 +188,6 @@ static const struct iio_chan_spec ad7124_channel_template = {
 		.sign = 'u',
 		.realbits = 24,
 		.storagebits = 32,
-		.shift = 8,
 		.endianness = IIO_BE,
 	},
 };
Signature · 1 line
2.35.1
author maintainer
#4 lore raw

PATCH 4/298 lkdtm/bugs: Check for the NULL pointer after calling kmalloc

In reply to Greg Kroah-Hartman

From: Jiasheng Jiang <[email protected]>

[ Upstream commit 4a9800c81d2f34afb66b4b42e0330ae8298019a2 ]

As the possible failure of the kmalloc(), the not_checked and checked could be NULL pointer.
Therefore, it should be better to check it in order to avoid the dereference of the NULL pointer.
Also, we need to kfree the 'not_checked' and 'checked' to avoid the memory leak if fails.
And since it is just a test, it may directly return without error number.

Fixes ae2e1aad3e48 ("drivers/misc/lkdtm/bugs.c: add arithmetic overflow and array bounds checks")
---
 drivers/misc/lkdtm/bugs.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/misc/lkdtm/bugs.c +5 -0
index f21854ac5cc2..4f2808b2ca3c 100644
@@ -327,6 +327,11 @@ void lkdtm_ARRAY_BOUNDS(void)
 
 	not_checked = kmalloc(sizeof(*not_checked) * 2, GFP_KERNEL);
 	checked = kmalloc(sizeof(*checked) * 2, GFP_KERNEL);
+	if (!not_checked || !checked) {
+		kfree(not_checked);
+		kfree(checked);
+		return;
+	}
 
 	pr_info("Array access within bounds ...\n");
 	/* For both, touch all bytes in the actual member size. */
Signature · 1 line
2.35.1
author maintainer
#5 lore raw

PATCH 5/298 lkdtm/bugs: Dont expect thread termination without CONFIG_UBSAN_TRAP

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

[ Upstream commit 8bfdbddd68249e0d8598777cca8249619ee51df0 ]

When you don't select CONFIG_UBSAN_TRAP, you get:

  # echo ARRAY_BOUNDS > /sys/kernel/debug/provoke-crash/DIRECT
[  102.265827] ================================================================================
[  102.278433] UBSAN: array-index-out-of-bounds in drivers/misc/lkdtm/bugs.c:342:16
[  102.287207] index 8 is out of range for type 'char [8]'
[  102.298722] ================================================================================
[  102.313712] lkdtm: FAIL: survived array bounds overflow!
[  102.318770] lkdtm: Unexpected! This kernel (5.16.0-rc1-s3k-dev-01884-g720dcf79314a ppc) was built with CONFIG_UBSAN_BOUNDS=y

It is not correct because when CONFIG_UBSAN_TRAP is not selected you can't expect array bounds overflow to kill the thread.

Modify the logic so that when the kernel is built with CONFIG_UBSAN_BOUNDS but without CONFIG_UBSAN_TRAP, you get a warning about CONFIG_UBSAN_TRAP not been selected instead.

This also require a fix of pr_expected_config(), otherwise the following error is encountered.

CC drivers/misc/lkdtm/bugs.o
drivers/misc/lkdtm/bugs.c: In function 'lkdtm_ARRAY_BOUNDS':
drivers/misc/lkdtm/bugs.c:351:2: error: 'else' without a previous 'if'

  351 |  else
      |  ^~~~
---
 drivers/misc/lkdtm/bugs.c  | 5 ++++-
 drivers/misc/lkdtm/lkdtm.h | 8 ++++----
 2 files changed, 8 insertions(+), 5 deletions(-)
drivers/misc/lkdtm/bugs.c +4 -1
index 4f2808b2ca3c..8cb342c562af 100644
@@ -351,7 +351,10 @@ void lkdtm_ARRAY_BOUNDS(void)
 	kfree(not_checked);
 	kfree(checked);
 	pr_err("FAIL: survived array bounds overflow!\n");
-	pr_expected_config(CONFIG_UBSAN_BOUNDS);
+	if (IS_ENABLED(CONFIG_UBSAN_BOUNDS))
+		pr_expected_config(CONFIG_UBSAN_TRAP);
+	else
+		pr_expected_config(CONFIG_UBSAN_BOUNDS);
 }
 
 void lkdtm_CORRUPT_LIST_ADD(void)
drivers/misc/lkdtm/lkdtm.h +4 -4
index d6137c70ebbe..cc76ebcca4c7 100644
@@ -9,19 +9,19 @@
 extern char *lkdtm_kernel_info;
 
 #define pr_expected_config(kconfig)				\
-{								\
+do {								\
 	if (IS_ENABLED(kconfig)) 				\
 		pr_err("Unexpected! This %s was built with " #kconfig "=y\n", \
 			lkdtm_kernel_info);			\
 	else							\
 		pr_warn("This is probably expected, since this %s was built *without* " #kconfig "=y\n", \
 			lkdtm_kernel_info);			\
-}
+} while (0)
 
 #ifndef MODULE
 int lkdtm_check_bool_cmdline(const char *param);
 #define pr_expected_config_param(kconfig, param)		\
-{								\
+do {								\
 	if (IS_ENABLED(kconfig)) {				\
 		switch (lkdtm_check_bool_cmdline(param)) {	\
 		case 0:						\
@@ -52,7 +52,7 @@ int lkdtm_check_bool_cmdline(const char *param);
 			break;					\
 		}						\
 	}							\
-}
+} while (0)
 #else
 #define pr_expected_config_param(kconfig, param) pr_expected_config(kconfig)
 #endif
Signature · 1 line
2.35.1
author maintainer
#6 lore raw

PATCH 6/298 tty: goldfish: Use tty_port_destroy() to destroy port

In reply to Greg Kroah-Hartman

From: Wang Weiyang <[email protected]>

[ Upstream commit 507b05063d1b7a1fcb9f7d7c47586fc4f3508f98 ]

In goldfish_tty_probe(), the port initialized through tty_port_init() should be destroyed in error paths.In goldfish_tty_remove(), qtty->port also should be destroyed or else might leak resources.

Fix the above by calling tty_port_destroy().

---
 drivers/tty/goldfish.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/tty/goldfish.c +2 -0
index 10c13b93ed52..9355d97ff591 100644
@@ -405,6 +405,7 @@ static int goldfish_tty_probe(struct platform_device *pdev)
 err_tty_register_device_failed:
 	free_irq(irq, qtty);
 err_dec_line_count:
+	tty_port_destroy(&qtty->port);
 	goldfish_tty_current_line_count--;
 	if (goldfish_tty_current_line_count == 0)
 		goldfish_tty_delete_driver();
@@ -426,6 +427,7 @@ static int goldfish_tty_remove(struct platform_device *pdev)
 	iounmap(qtty->base);
 	qtty->base = NULL;
 	free_irq(qtty->irq, pdev);
+	tty_port_destroy(&qtty->port);
 	goldfish_tty_current_line_count--;
 	if (goldfish_tty_current_line_count == 0)
 		goldfish_tty_delete_driver();
Signature · 1 line
2.35.1
author maintainer
#7 lore raw

PATCH 7/298 tty: serial: owl: Fix missing clk_disable_unprepare() in owl_uart_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit bcea0f547ec1a2ee44d429aaf0334633e386e67c ]

Fix the missing clk_disable_unprepare() before return from owl_uart_probe() in the error handling case.

Fixes abf42d2f333b ("tty: serial: owl: add "much needed" clk_prepare_enable()")
---
 drivers/tty/serial/owl-uart.c | 1 +
 1 file changed, 1 insertion(+)
drivers/tty/serial/owl-uart.c +1 -0
index 91f1eb0058d7..9a6611cfc18e 100644
@@ -731,6 +731,7 @@ static int owl_uart_probe(struct platform_device *pdev)
 	owl_port->port.uartclk = clk_get_rate(owl_port->clk);
 	if (owl_port->port.uartclk == 0) {
 		dev_err(&pdev->dev, "clock rate is zero\n");
+		clk_disable_unprepare(owl_port->clk);
 		return -EINVAL;
 	}
 	owl_port->port.flags = UPF_BOOT_AUTOCONF | UPF_IOREMAP | UPF_LOW_LATENCY;
Signature · 1 line
2.35.1
author maintainer
#8 lore raw

PATCH 8/298 tty: n_tty: Restore EOF push handling behavior

In reply to Greg Kroah-Hartman

From: Daniel Gibson <[email protected]>

[ Upstream commit 65a8b287023da68c4550deab5c764e6891cf1caf ]

TTYs in ICANON mode have a special case that allows "pushing" a line without a regular EOL character (like newline), by using EOF (the EOT character - ASCII 0x4) as a pseudo-EOL. It is silently discarded, so the reader of the PTS will receive the line *without* EOF or any other terminating character.

This special case has an edge case: What happens if the readers buffer is the same size as the line (without EOF)? Will they be able to tell if the whole line is received, i.e. if the next read() will return more of the same line or the next line?

There are two possibilities, that both have (dis)advantages:

1. The next read() returns 0. FreeBSD (13.0) and OSX (10.11) do this.
Advantage: The reader can interpret this as "the line is over".
Disadvantage: read() returning 0 means EOF, the reader could also interpret it as "there's no more data" and stop reading or even close the PT.

2. The next read() returns the next line, the EOF is silently discarded. Solaris (or at least OpenIndiana 2021.10) does this, Linux has done do this since commit 40d5e0905a03 ("n_tty: Fix EOF push handling"); this behavior was recently broken by commit 359303076163 ("tty:
n_tty: do not look ahead for EOL character past the end of the buffer").
Advantage: read() won't return 0 (EOF), reader less likely to be confused (and things like while(read(..)>0) don't break)
Disadvantage: The reader can't really know if the read() continues the last line (that filled the whole read buffer) or starts a new line.

As both options are defensible (and are used by other Unix-likes), it's best to stick to the "old" behavior since "n_tty: Fix EOF push handling" of 2013, i.e. silently discard that EOF.

This patch - that I actually got from Linus for testing and only modified slightly - restores that behavior by skipping an EOF character if it's the next character after reading is done.

Based on a patch from Linus Torvalds.

Fixes 359303076163 ("tty: n_tty: do not look ahead for EOL character past the end of the buffer")
Cc Peter Hurley <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc Jiri Slaby <[email protected]>
Reviewed-and-tested-by Daniel Gibson <[email protected]>
---
 drivers/tty/n_tty.c | 38 +++++++++++++++++++++++++++++++++++++-
 1 file changed, 37 insertions(+), 1 deletion(-)
drivers/tty/n_tty.c +37 -1
index efc72104c840..bdc314aeab88 100644
@@ -1975,6 +1975,35 @@ static bool canon_copy_from_read_buf(struct tty_struct *tty,
 	return ldata->read_tail != canon_head;
 }
 
+/*
+ * If we finished a read at the exact location of an
+ * EOF (special EOL character that's a __DISABLED_CHAR)
+ * in the stream, silently eat the EOF.
+ */
+static void canon_skip_eof(struct tty_struct *tty)
+{
+	struct n_tty_data *ldata = tty->disc_data;
+	size_t tail, canon_head;
+
+	canon_head = smp_load_acquire(&ldata->canon_head);
+	tail = ldata->read_tail;
+
+	// No data?
+	if (tail == canon_head)
+		return;
+
+	// See if the tail position is EOF in the circular buffer
+	tail &= (N_TTY_BUF_SIZE - 1);
+	if (!test_bit(tail, ldata->read_flags))
+		return;
+	if (read_buf(ldata, tail) != __DISABLED_CHAR)
+		return;
+
+	// Clear the EOL bit, skip the EOF char.
+	clear_bit(tail, ldata->read_flags);
+	smp_store_release(&ldata->read_tail, ldata->read_tail + 1);
+}
+
 /**
  * job_control		-	check job control
  * @tty: tty
@@ -2045,7 +2074,14 @@ static ssize_t n_tty_read(struct tty_struct *tty, struct file *file,
 	 */
 	if (*cookie) {
 		if (ldata->icanon && !L_EXTPROC(tty)) {
-			if (canon_copy_from_read_buf(tty, &kb, &nr))
+			/*
+			 * If we have filled the user buffer, see
+			 * if we should skip an EOF character before
+			 * releasing the lock and returning done.
+			 */
+			if (!nr)
+				canon_skip_eof(tty);
+			else if (canon_copy_from_read_buf(tty, &kb, &nr))
 				return kb - kbuf;
 		} else {
 			if (copy_from_read_buf(tty, &kb, &nr))
Signature · 1 line
2.35.1
author maintainer
#9 lore raw

PATCH 9/298 serial: 8250_aspeed_vuart: Fix potential NULL dereference in aspeed_vuart_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 0e0fd55719fa081de6f9e5d9e6cef48efb04d34a ]

platform_get_resource() may fail and return NULL, so we should better check it's return value to avoid a NULL pointer dereference.

Fixes 54da3e381c2b ("serial: 8250_aspeed_vuart: use UPF_IOREMAP to set up register mapping")
---
 drivers/tty/serial/8250/8250_aspeed_vuart.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/tty/serial/8250/8250_aspeed_vuart.c +2 -0
index c2cecc6f47db..179bb1375636 100644
@@ -429,6 +429,8 @@ static int aspeed_vuart_probe(struct platform_device *pdev)
 	timer_setup(&vuart->unthrottle_timer, aspeed_vuart_unthrottle_exp, 0);
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -EINVAL;
 
 	memset(&port, 0, sizeof(port));
 	port.port.private_data = vuart;
Signature · 1 line
2.35.1
author maintainer
#10 lore raw

PATCH 10/298 tty: serial: fsl_lpuart: fix potential bug when using both of_alias_get_id and ida_simple_get

In reply to Greg Kroah-Hartman

From: Sherry Sun <[email protected]>

[ Upstream commit f398e0aa325c61fa20903833a5b534ecb8e6e418 ]

Now fsl_lpuart driver use both of_alias_get_id() and ida_simple_get() in .probe(), which has the potential bug. For example, when remove the lpuart7 alias in dts, of_alias_get_id() will return error, then call ida_simple_get() to allocate the id 0 for lpuart7, this may confilct with the lpuart4 which has alias 0.

    aliases {
	...
        serial0 = &lpuart4;
        serial1 = &lpuart5;
        serial2 = &lpuart6;
        serial3 = &lpuart7;
    }

So remove the ida_simple_get() in .probe(), return an error directly when calling of_alias_get_id() fails, which is consistent with other uart drivers behavior.

Fixes 3bc3206e1c0f ("serial: fsl_lpuart: Remove the alias node dependence")
---
 drivers/tty/serial/fsl_lpuart.c | 24 ++++--------------------
 1 file changed, 4 insertions(+), 20 deletions(-)
drivers/tty/serial/fsl_lpuart.c +4 -20
index d32c25bc973b..b1307ef34468 100644
@@ -239,8 +239,6 @@
 /* IMX lpuart has four extra unused regs located at the beginning */
 #define IMX_REG_OFF	0x10
 
-static DEFINE_IDA(fsl_lpuart_ida);
-
 enum lpuart_type {
 	VF610_LPUART,
 	LS1021A_LPUART,
@@ -276,7 +274,6 @@ struct lpuart_port {
 	int			rx_dma_rng_buf_len;
 	unsigned int		dma_tx_nents;
 	wait_queue_head_t	dma_wait;
-	bool			id_allocated;
 };
 
 struct lpuart_soc_data {
@@ -2711,23 +2708,18 @@ static int lpuart_probe(struct platform_device *pdev)
 
 	ret = of_alias_get_id(np, "serial");
 	if (ret < 0) {
-		ret = ida_simple_get(&fsl_lpuart_ida, 0, UART_NR, GFP_KERNEL);
-		if (ret < 0) {
-			dev_err(&pdev->dev, "port line is full, add device failed\n");
-			return ret;
-		}
-		sport->id_allocated = true;
+		dev_err(&pdev->dev, "failed to get alias id, errno %d\n", ret);
+		return ret;
 	}
 	if (ret >= ARRAY_SIZE(lpuart_ports)) {
 		dev_err(&pdev->dev, "serial%d out of range\n", ret);
-		ret = -EINVAL;
-		goto failed_out_of_range;
+		return -EINVAL;
 	}
 	sport->port.line = ret;
 
 	ret = lpuart_enable_clks(sport);
 	if (ret)
-		goto failed_clock_enable;
+		return ret;
 	sport->port.uartclk = lpuart_get_baud_clk_rate(sport);
 
 	lpuart_ports[sport->port.line] = sport;
@@ -2775,10 +2767,6 @@ static int lpuart_probe(struct platform_device *pdev)
 	uart_remove_one_port(&lpuart_reg, &sport->port);
 failed_attach_port:
 	lpuart_disable_clks(sport);
-failed_clock_enable:
-failed_out_of_range:
-	if (sport->id_allocated)
-		ida_simple_remove(&fsl_lpuart_ida, sport->port.line);
 	return ret;
 }
 
@@ -2788,9 +2776,6 @@ static int lpuart_remove(struct platform_device *pdev)
 
 	uart_remove_one_port(&lpuart_reg, &sport->port);
 
-	if (sport->id_allocated)
-		ida_simple_remove(&fsl_lpuart_ida, sport->port.line);
-
 	lpuart_disable_clks(sport);
 
 	if (sport->dma_tx_chan)
@@ -2920,7 +2905,6 @@ static int __init lpuart_serial_init(void)
 
 static void __exit lpuart_serial_exit(void)
 {
-	ida_destroy(&fsl_lpuart_ida);
 	platform_driver_unregister(&lpuart_driver);
 	uart_unregister_driver(&lpuart_reg);
 }
Signature · 1 line
2.35.1
author maintainer
#11 lore raw

PATCH 11/298 remoteproc: imx_rproc: Ignore create mem entry for resource table

In reply to Greg Kroah-Hartman

From: Peng Fan <[email protected]>

[ Upstream commit 58b7c856519fe946620ee68dd0c37bd3c695484a ]

Resource table is used by Linux to get information published by remote processor. It should be not be used for memory allocation, so not create rproc mem entry.

Fixes b29b4249f8f0 ("remoteproc: imx_rproc: add i.MX specific parse fw hook")
---
 drivers/remoteproc/imx_rproc.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/remoteproc/imx_rproc.c +3 -0
index 7a096f1891e6..91eb037089ef 100644
@@ -423,6 +423,9 @@ static int imx_rproc_prepare(struct rproc *rproc)
 		if (!strcmp(it.node->name, "vdev0buffer"))
 			continue;
 
+		if (!strcmp(it.node->name, "rsc-table"))
+			continue;
+
 		rmem = of_reserved_mem_lookup(it.node);
 		if (!rmem) {
 			dev_err(priv->dev, "unable to acquire memory-region\n");
Signature · 1 line
2.35.1
author maintainer
#12 lore raw

PATCH 12/298 phy: rockchip-inno-usb2: Fix muxed interrupt support

In reply to Greg Kroah-Hartman

From: Samuel Holland <[email protected]>

[ Upstream commit 6a98df08ccd55e87947d253b19925691763e755c ]

This commit fixes two issues with the muxed interrupt handler. First, the OTG port has the "bvalid" interrupt enabled, not "linestate". Since only the linestate interrupt was handled, and not the bvalid interrupt, plugging in a cable to the OTG port caused an interrupt storm.

Second, the return values from the individual port IRQ handlers need to be OR-ed together. Otherwise, the lack of an interrupt from the last port would cause the handler to erroneously return IRQ_NONE.

Fixes ed2b5a8e6b98 ("phy: phy-rockchip-inno-usb2: support muxed interrupts")
---
 drivers/phy/rockchip/phy-rockchip-inno-usb2.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/phy/rockchip/phy-rockchip-inno-usb2.c +8 -2
index eca77e44a4c1..cba5c32cbaee 100644
@@ -940,8 +940,14 @@ static irqreturn_t rockchip_usb2phy_irq(int irq, void *data)
 		if (!rport->phy)
 			continue;
 
-		/* Handle linestate irq for both otg port and host port */
-		ret = rockchip_usb2phy_linestate_irq(irq, rport);
+		switch (rport->port_id) {
+		case USB2PHY_PORT_OTG:
+			ret |= rockchip_usb2phy_otg_mux_irq(irq, rport);
+			break;
+		case USB2PHY_PORT_HOST:
+			ret |= rockchip_usb2phy_linestate_irq(irq, rport);
+			break;
+		}
 	}
 
 	return ret;
Signature · 1 line
2.35.1
author maintainer
#13 lore raw

PATCH 13/298 usb: usbip: fix a refcount leak in stub_probe()

In reply to Greg Kroah-Hartman

From: Hangyu Hua <[email protected]>

[ Upstream commit 9ec4cbf1cc55d126759051acfe328d489c5d6e60 ]

usb_get_dev() is called in stub_device_alloc(). When stub_probe() fails after that, usb_put_dev() needs to be called to release the reference.

Fix this by moving usb_put_dev() to sdev_free error path handling.

Find this by code review.

Fixes 3ff67445750a ("usbip: fix error handling in stub_probe()")
---
 drivers/usb/usbip/stub_dev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/usb/usbip/stub_dev.c +1 -1
index d8d3892e5a69..3c6d452e3bf4 100644
@@ -393,7 +393,6 @@ static int stub_probe(struct usb_device *udev)
 
 err_port:
 	dev_set_drvdata(&udev->dev, NULL);
-	usb_put_dev(udev);
 
 	/* we already have busid_priv, just lock busid_lock */
 	spin_lock(&busid_priv->busid_lock);
@@ -408,6 +407,7 @@ static int stub_probe(struct usb_device *udev)
 	put_busid_priv(busid_priv);
 
 sdev_free:
+	usb_put_dev(udev);
 	stub_device_free(sdev);
 
 	return rc;
Signature · 1 line
2.35.1
author maintainer
#14 lore raw

PATCH 14/298 usb: usbip: add missing device lock on tweak configuration cmd

In reply to Greg Kroah-Hartman

From: Niels Dossche <[email protected]>

[ Upstream commit d088fabace2ca337b275d1d4b36db4fe7771e44f ]

The function documentation of usb_set_configuration says that its callers should hold the device lock. This lock is held for all callsites except tweak_set_configuration_cmd. The code path can be executed for example when attaching a remote USB device. The solution is to surround the call by the device lock.

This bug was found using my experimental own-developed static analysis tool, which reported the missing lock on v5.17.2. I manually verified this bug report by doing code review as well. I runtime checked that the required lock is not held. I compiled and runtime tested this on x86_64 with a USB mouse. After applying this patch, my analyser no longer reports this potential bug.

Fixes 2c8c98158946 ("staging: usbip: let client choose device configuration")
---
 drivers/usb/usbip/stub_rx.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/usb/usbip/stub_rx.c +2 -0
index 325c22008e53..5dd41e8215e0 100644
@@ -138,7 +138,9 @@ static int tweak_set_configuration_cmd(struct urb *urb)
 	req = (struct usb_ctrlrequest *) urb->setup_packet;
 	config = le16_to_cpu(req->wValue);
 
+	usb_lock_device(sdev->udev);
 	err = usb_set_configuration(sdev->udev, config);
+	usb_unlock_device(sdev->udev);
 	if (err && err != -ENODEV)
 		dev_err(&sdev->udev->dev, "can't set config #%d, error %d\n",
 			config, err);
Signature · 1 line
2.35.1
author maintainer
#15 lore raw

PATCH 15/298 USB: storage: karma: fix rio_karma_init return

In reply to Greg Kroah-Hartman

From: Lin Ma <[email protected]>

[ Upstream commit b92ffb1eddd9a66a90defc556dcbf65a43c196c7 ]

The function rio_karam_init() should return -ENOMEM instead of value 0 (USB_STOR_TRANSPORT_GOOD) when allocation fails.

Similarly, it should return -EIO when rio_karma_send_command() fails.

---
 drivers/usb/storage/karma.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
drivers/usb/storage/karma.c +8 -7
index 05cec81dcd3f..38ddfedef629 100644
@@ -174,24 +174,25 @@ static void rio_karma_destructor(void *extra)
 
 static int rio_karma_init(struct us_data *us)
 {
-	int ret = 0;
 	struct karma_data *data = kzalloc(sizeof(struct karma_data), GFP_NOIO);
 
 	if (!data)
-		goto out;
+		return -ENOMEM;
 
 	data->recv = kmalloc(RIO_RECV_LEN, GFP_NOIO);
 	if (!data->recv) {
 		kfree(data);
-		goto out;
+		return -ENOMEM;
 	}
 
 	us->extra = data;
 	us->extra_destructor = rio_karma_destructor;
-	ret = rio_karma_send_command(RIO_ENTER_STORAGE, us);
-	data->in_storage = (ret == 0);
-out:
-	return ret;
+	if (rio_karma_send_command(RIO_ENTER_STORAGE, us))
+		return -EIO;
+
+	data->in_storage = 1;
+
+	return 0;
 }
 
 static struct scsi_host_template karma_host_template;
Signature · 1 line
2.35.1
author maintainer
#16 lore raw

PATCH 16/298 usb: musb: Fix missing of_node_put() in omap2430_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 424bef51fa530389b0b9008c9e144e40c10e8458 ]

The device_node pointer is returned by of_parse_phandle() with refcount incremented. We should use of_node_put() on it when done.

Fixes 8934d3e4d0e7 ("usb: musb: omap2430: Don't use omap_get_control_dev()")
---
 drivers/usb/musb/omap2430.c | 1 +
 1 file changed, 1 insertion(+)
drivers/usb/musb/omap2430.c +1 -0
index d2b7e613eb34..f571a65ae6ee 100644
@@ -362,6 +362,7 @@ static int omap2430_probe(struct platform_device *pdev)
 	control_node = of_parse_phandle(np, "ctrl-module", 0);
 	if (control_node) {
 		control_pdev = of_find_device_by_node(control_node);
+		of_node_put(control_node);
 		if (!control_pdev) {
 			dev_err(&pdev->dev, "Failed to get control device\n");
 			ret = -EINVAL;
Signature · 1 line
2.35.1
author maintainer
#17 lore raw

PATCH 17/298 staging: fieldbus: Fix the error handling path in anybuss_host_common_probe()

In reply to Greg Kroah-Hartman

From: Christophe JAILLET <[email protected]>

[ Upstream commit 7079b3483a17be2cfba64cbd4feb1b7ae07f1ea7 ]

If device_register() fails, device_unregister() should not be called because it will free some resources that are not allocated. put_device() should be used instead.

---
 drivers/staging/fieldbus/anybuss/host.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/fieldbus/anybuss/host.c +1 -1
index a344410e48fe..cd86b9c9e345 100644
@@ -1384,7 +1384,7 @@ anybuss_host_common_probe(struct device *dev,
 		goto err_device;
 	return cd;
 err_device:
-	device_unregister(&cd->client->dev);
+	put_device(&cd->client->dev);
 err_kthread:
 	kthread_stop(cd->qthread);
 err_reset:
Signature · 1 line
2.35.1
author maintainer
#18 lore raw

PATCH 18/298 pwm: lp3943: Fix duty calculation in case period was clamped

In reply to Greg Kroah-Hartman

From: Uwe Kleine-König <[email protected]>

[ Upstream commit 5e3b07ca5cc78cd4a987e78446849e41288d87cb ]

The hardware only supports periods <= 1.6 ms and if a bigger period is requested it is clamped to 1.6 ms. In this case duty_cycle might be bigger than 1.6 ms and then the duty cycle register is written with a value bigger than LP3943_MAX_DUTY. So clamp duty_cycle accordingly.

Fixes af66b3c0934e ("pwm: Add LP3943 PWM driver")
---
 drivers/pwm/pwm-lp3943.c | 1 +
 1 file changed, 1 insertion(+)
drivers/pwm/pwm-lp3943.c +1 -0
index ea17d446a627..2bd04ecb508c 100644
@@ -125,6 +125,7 @@ static int lp3943_pwm_config(struct pwm_chip *chip, struct pwm_device *pwm,
 	if (err)
 		return err;
 
+	duty_ns = min(duty_ns, period_ns);
 	val = (u8)(duty_ns * LP3943_MAX_DUTY / period_ns);
 
 	return lp3943_write_byte(lp3943, reg_duty, val);
Signature · 1 line
2.35.1
author maintainer
#19 lore raw

PATCH 19/298 pwm: raspberrypi-poe: Fix endianness in firmware struct

In reply to Greg Kroah-Hartman

From: Uwe Kleine-König <[email protected]>

[ Upstream commit 09f688f0718f57f9cf68ee1aa94490f641e759ba ]

The reg member of struct raspberrypi_pwm_prop is a little endian 32 bit quantity. Explicitly convert the (native endian) value to little endian on assignment as is already done in raspberrypi_pwm_set_property().

This fixes the following sparse warning:

drivers/pwm/pwm-raspberrypi-poe.c:69:24: warning: incorrect type in initializer (different base types)
drivers/pwm/pwm-raspberrypi-poe.c:69:24: expected restricted __le32 [usertype] reg
drivers/pwm/pwm-raspberrypi-poe.c:69:24: got unsigned int [usertype] reg

Fixes 79caa362eab6 ("pwm: Add Raspberry Pi Firmware based PWM bus")
---
 drivers/pwm/pwm-raspberrypi-poe.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pwm/pwm-raspberrypi-poe.c +1 -1
index 579a15240e0a..c877de37734d 100644
@@ -66,7 +66,7 @@ static int raspberrypi_pwm_get_property(struct rpi_firmware *firmware,
 					u32 reg, u32 *val)
 {
 	struct raspberrypi_pwm_prop msg = {
-		.reg = reg
+		.reg = cpu_to_le32(reg),
 	};
 	int ret;
 
Signature · 1 line
2.35.1
author maintainer
#20 lore raw

PATCH 20/298 rpmsg: qcom_smd: Fix irq_of_parse_and_map() return value

In reply to Greg Kroah-Hartman

From: Krzysztof Kozlowski <[email protected]>

[ Upstream commit 1a358d35066487d228a68303d808bc4721c6b1b9 ]

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

Fixes 53e2822e56c7 ("rpmsg: Introduce Qualcomm SMD backend")
---
 drivers/rpmsg/qcom_smd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rpmsg/qcom_smd.c +1 -1
index 540e027f08c4..d4b54eebe15d 100644
@@ -1404,7 +1404,7 @@ static int qcom_smd_parse_edge(struct device *dev,
 		edge->name = node->name;
 
 	irq = irq_of_parse_and_map(node, 0);
-	if (irq < 0) {
+	if (!irq) {
 		dev_err(dev, "required smd interrupt missing\n");
 		ret = irq;
 		goto put_node;
Signature · 1 line
2.35.1
author maintainer
#21 lore raw

PATCH 21/298 usb: dwc3: gadget: Replace list_for_each_entry_safe() if using giveback

In reply to Greg Kroah-Hartman

From: Wesley Cheng <[email protected]>

[ Upstream commit bf594d1d0c1d7b895954018043536ffd327844f9 ]

The list_for_each_entry_safe() macro saves the current item (n) and the item after (n+1), so that n can be safely removed without corrupting the list. However, when traversing the list and removing items using gadget giveback, the DWC3 lock is briefly released, allowing other routines to execute. There is a situation where, while items are being removed from the cancelled_list using dwc3_gadget_ep_cleanup_cancelled_requests(), the pullup disable routine is running in parallel (due to UDC unbind). As the cleanup routine removes n, and the pullup disable removes n+1, once the cleanup retakes the DWC3 lock, it references a request who was already removed/handled. With list debug enabled, this leads to a panic. Ensure all instances of the macro are replaced where gadget giveback is used.

Example call stack:

Thread#1:
__dwc3_gadget_ep_set_halt() - CLEAR HALT
-> dwc3_gadget_ep_cleanup_cancelled_requests()

    ->list_for_each_entry_safe()
    ->dwc3_gadget_giveback(n)
      ->dwc3_gadget_del_and_unmap_request()- n deleted[cancelled_list]
      ->spin_unlock
      ->Thread#2 executes
      ...
    ->dwc3_gadget_giveback(n+1)
      ->Already removed!

Thread#2:
dwc3_gadget_pullup()
->waiting for dwc3 spin_lock
...
->Thread#1 released lock
->dwc3_stop_active_transfers()

    ->dwc3_remove_requests()
      ->fetches n+1 item from cancelled_list (n removed by Thread#1)
      ->dwc3_gadget_giveback()
        ->dwc3_gadget_del_and_unmap_request()- n+1 deleted[cancelled_list]
        ->spin_unlock
Fixes d4f1afe5e896 ("usb: dwc3: gadget: move requests to cancelled_list")
---
 drivers/usb/dwc3/gadget.c | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)
drivers/usb/dwc3/gadget.c +16 -4
index 78ec6af79c7f..5c1ae0d0ed47 100644
@@ -1978,10 +1978,10 @@ static void dwc3_gadget_ep_skip_trbs(struct dwc3_ep *dep, struct dwc3_request *r
 static void dwc3_gadget_ep_cleanup_cancelled_requests(struct dwc3_ep *dep)
 {
 	struct dwc3_request		*req;
-	struct dwc3_request		*tmp;
 	struct dwc3			*dwc = dep->dwc;
 
-	list_for_each_entry_safe(req, tmp, &dep->cancelled_list, list) {
+	while (!list_empty(&dep->cancelled_list)) {
+		req = next_request(&dep->cancelled_list);
 		dwc3_gadget_ep_skip_trbs(dep, req);
 		switch (req->status) {
 		case DWC3_REQUEST_STATUS_DISCONNECTED:
@@ -1998,6 +1998,12 @@ static void dwc3_gadget_ep_cleanup_cancelled_requests(struct dwc3_ep *dep)
 			dwc3_gadget_giveback(dep, req, -ECONNRESET);
 			break;
 		}
+		/*
+		 * The endpoint is disabled, let the dwc3_remove_requests()
+		 * handle the cleanup.
+		 */
+		if (!dep->endpoint.desc)
+			break;
 	}
 }
 
@@ -3288,15 +3294,21 @@ static void dwc3_gadget_ep_cleanup_completed_requests(struct dwc3_ep *dep,
 		const struct dwc3_event_depevt *event, int status)
 {
 	struct dwc3_request	*req;
-	struct dwc3_request	*tmp;
 
-	list_for_each_entry_safe(req, tmp, &dep->started_list, list) {
+	while (!list_empty(&dep->started_list)) {
 		int ret;
 
+		req = next_request(&dep->started_list);
 		ret = dwc3_gadget_ep_cleanup_completed_request(dep, event,
 				req, status);
 		if (ret)
 			break;
+		/*
+		 * The endpoint is disabled, let the dwc3_remove_requests()
+		 * handle the cleanup.
+		 */
+		if (!dep->endpoint.desc)
+			break;
 	}
 }
 
Signature · 1 line
2.35.1
author maintainer
#22 lore raw

PATCH 22/298 usb: dwc3: pci: Fix pm_runtime_get_sync() error checking

In reply to Greg Kroah-Hartman

From: Zheng Yongjun <[email protected]>

[ Upstream commit a03e2ddab8e735e2cc315609b297b300e9cc60d2 ]

If the device is already in a runtime PM enabled state pm_runtime_get_sync() will return 1, so a test for negative value should be used to check for errors.

Fixes 8eed00b237a28 ("usb: dwc3: pci: Runtime resume child device from wq")
---
 drivers/usb/dwc3/dwc3-pci.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/usb/dwc3/dwc3-pci.c +1 -1
index f08b2178fd32..9c8887615701 100644
@@ -256,7 +256,7 @@ static void dwc3_pci_resume_work(struct work_struct *work)
 	int ret;
 
 	ret = pm_runtime_get_sync(&dwc3->dev);
-	if (ret) {
+	if (ret < 0) {
 		pm_runtime_put_sync_autosuspend(&dwc3->dev);
 		return;
 	}
Signature · 1 line
2.35.1
author maintainer
#23 lore raw

PATCH 23/298 scripts/get_abi: Fix wrong script file name in the help message

In reply to Greg Kroah-Hartman

From: SeongJae Park <[email protected]>

[ Upstream commit 5b5bfecaa333fb6a0cce1bfc4852a622dacfed1d ]

The help message of 'get_abi.pl' is mistakenly saying it's 'abi_book.pl'. This commit fixes the wrong name in the help message.

Fixes bbc249f2b859 ("scripts: add an script to parse the ABI files")
---
 scripts/get_abi.pl | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
scripts/get_abi.pl +2 -2
index 6212f58b69c6..0cf501285204 100755
@@ -980,11 +980,11 @@ __END__
 
 =head1 NAME
 
-abi_book.pl - parse the Linux ABI files and produce a ReST book.
+get_abi.pl - parse the Linux ABI files and produce a ReST book.
 
 =head1 SYNOPSIS
 
-B<abi_book.pl> [--debug <level>] [--enable-lineno] [--man] [--help]
+B<get_abi.pl> [--debug <level>] [--enable-lineno] [--man] [--help]
 	       [--(no-)rst-source] [--dir=<dir>] [--show-hints]
 	       [--search-string <regex>]
 	       <COMAND> [<ARGUMENT>]
Signature · 1 line
2.35.1
author maintainer
#24 lore raw

PATCH 24/298 misc: fastrpc: fix an incorrect NULL check on list iterator

In reply to Greg Kroah-Hartman

From: Xiaomeng Tong <[email protected]>

[ Upstream commit 5ac11fe03a0a83042d1a040dbce4fa2fb5521e23 ]

The bug is here:

	if (!buf) {

The list iterator value 'buf' will *always* be set and non-NULL by list_for_each_entry(), so it is incorrect to assume that the iterator value will be NULL if the list is empty (in this case, the check 'if (!buf) {' will always be false and never exit expectly).

To fix the bug, use a new variable 'iter' as the list iterator, while use the original variable 'buf' as a dedicated pointer to point to the found element.

Fixes 2419e55e532de ("misc: fastrpc: add mmap/unmap support")
---
 drivers/misc/fastrpc.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/misc/fastrpc.c +5 -4
index aa1682b94a23..45aaf54a7560 100644
@@ -1353,17 +1353,18 @@ static int fastrpc_req_munmap_impl(struct fastrpc_user *fl,
 				   struct fastrpc_req_munmap *req)
 {
 	struct fastrpc_invoke_args args[1] = { [0] = { 0 } };
-	struct fastrpc_buf *buf, *b;
+	struct fastrpc_buf *buf = NULL, *iter, *b;
 	struct fastrpc_munmap_req_msg req_msg;
 	struct device *dev = fl->sctx->dev;
 	int err;
 	u32 sc;
 
 	spin_lock(&fl->lock);
-	list_for_each_entry_safe(buf, b, &fl->mmaps, node) {
-		if ((buf->raddr == req->vaddrout) && (buf->size == req->size))
+	list_for_each_entry_safe(iter, b, &fl->mmaps, node) {
+		if ((iter->raddr == req->vaddrout) && (iter->size == req->size)) {
+			buf = iter;
 			break;
-		buf = NULL;
+		}
 	}
 	spin_unlock(&fl->lock);
 
Signature · 1 line
2.35.1
author maintainer
#25 lore raw

PATCH 25/298 firmware: stratix10-svc: fix a missing check on list iterator

In reply to Greg Kroah-Hartman

From: Xiaomeng Tong <[email protected]>

[ Upstream commit 5a0793ac66ac0e254d292f129a4d6c526f9f2aff ]

The bug is here:

	pmem->vaddr = NULL;

The list iterator 'pmem' will point to a bogus position containing HEAD if the list is empty or no element is found. This case must be checked before any use of the iterator, otherwise it will lead to a invalid memory access.

To fix this bug, just gen_pool_free/set NULL/list_del() and return when found, otherwise list_del HEAD and return;

Fixes 7ca5ce896524f ("firmware: add Intel Stratix10 service layer driver")
---
 drivers/firmware/stratix10-svc.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/firmware/stratix10-svc.c +6 -6
index c4bf934e3553..33ab24cfd600 100644
@@ -941,17 +941,17 @@ EXPORT_SYMBOL_GPL(stratix10_svc_allocate_memory);
 void stratix10_svc_free_memory(struct stratix10_svc_chan *chan, void *kaddr)
 {
 	struct stratix10_svc_data_mem *pmem;
-	size_t size = 0;
 
 	list_for_each_entry(pmem, &svc_data_mem, node)
 		if (pmem->vaddr == kaddr) {
-			size = pmem->size;
-			break;
+			gen_pool_free(chan->ctrl->genpool,
+				       (unsigned long)kaddr, pmem->size);
+			pmem->vaddr = NULL;
+			list_del(&pmem->node);
+			return;
 		}
 
-	gen_pool_free(chan->ctrl->genpool, (unsigned long)kaddr, size);
-	pmem->vaddr = NULL;
-	list_del(&pmem->node);
+	list_del(&svc_data_mem);
 }
 EXPORT_SYMBOL_GPL(stratix10_svc_free_memory);
 
Signature · 1 line
2.35.1
author maintainer
#26 lore raw

PATCH 26/298 usb: typec: mux: Check dev_set_name() return value

In reply to Greg Kroah-Hartman

From: Bjorn Andersson <[email protected]>

[ Upstream commit b9fa0292490db39d6542f514117333d366ec0011 ]

It's possible that dev_set_name() returns -ENOMEM, catch and handle this.

---
 drivers/usb/typec/mux.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
drivers/usb/typec/mux.c +10 -4
index c8340de0ed49..d2aaf294b649 100644
@@ -131,8 +131,11 @@ typec_switch_register(struct device *parent,
 	sw->dev.class = &typec_mux_class;
 	sw->dev.type = &typec_switch_dev_type;
 	sw->dev.driver_data = desc->drvdata;
-	dev_set_name(&sw->dev, "%s-switch",
-		     desc->name ? desc->name : dev_name(parent));
+	ret = dev_set_name(&sw->dev, "%s-switch", desc->name ? desc->name : dev_name(parent));
+	if (ret) {
+		put_device(&sw->dev);
+		return ERR_PTR(ret);
+	}
 
 	ret = device_add(&sw->dev);
 	if (ret) {
@@ -338,8 +341,11 @@ typec_mux_register(struct device *parent, const struct typec_mux_desc *desc)
 	mux->dev.class = &typec_mux_class;
 	mux->dev.type = &typec_mux_dev_type;
 	mux->dev.driver_data = desc->drvdata;
-	dev_set_name(&mux->dev, "%s-mux",
-		     desc->name ? desc->name : dev_name(parent));
+	ret = dev_set_name(&mux->dev, "%s-mux", desc->name ? desc->name : dev_name(parent));
+	if (ret) {
+		put_device(&mux->dev);
+		return ERR_PTR(ret);
+	}
 
 	ret = device_add(&mux->dev);
 	if (ret) {
Signature · 1 line
2.35.1
author maintainer
#27 lore raw

PATCH 27/298 rpmsg: virtio: Fix possible double free in rpmsg_probe()

In reply to Greg Kroah-Hartman

From: Hangyu Hua <[email protected]>

[ Upstream commit c2eecefec5df1306eafce28ccdf1ca159a552ecc ]

vch will be free in virtio_rpmsg_release_device() when rpmsg_ns_register_device() fails. There is no need to call kfree() again.

Fix this by changing error path from free_vch to free_ctrldev.

Fixes c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
---
 drivers/rpmsg/virtio_rpmsg_bus.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/rpmsg/virtio_rpmsg_bus.c +2 -3
index ac764e04c898..1c39e9c4fa02 100644
@@ -973,7 +973,8 @@ static int rpmsg_probe(struct virtio_device *vdev)
 
 		err = rpmsg_ns_register_device(rpdev_ns);
 		if (err)
-			goto free_vch;
+			/* vch will be free in virtio_rpmsg_release_device() */
+			goto free_ctrldev;
 	}
 
 	/*
@@ -997,8 +998,6 @@ static int rpmsg_probe(struct virtio_device *vdev)
 
 	return 0;
 
-free_vch:
-	kfree(vch);
 free_ctrldev:
 	rpmsg_virtio_del_ctrl_dev(rpdev_ctrl);
 free_coherent:
Signature · 1 line
2.35.1
author maintainer
#28 lore raw

PATCH 28/298 rpmsg: virtio: Fix possible double free in rpmsg_virtio_add_ctrl_dev()

In reply to Greg Kroah-Hartman

From: Hangyu Hua <[email protected]>

[ Upstream commit 1680939e9ecf7764fba8689cfb3429c2fe2bb23c ]

vch will be free in virtio_rpmsg_release_device() when rpmsg_ctrldev_register_device() fails. There is no need to call kfree() again.

Fixes c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
---
 drivers/rpmsg/virtio_rpmsg_bus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rpmsg/virtio_rpmsg_bus.c +1 -1
index 1c39e9c4fa02..99e0de65733a 100644
@@ -851,7 +851,7 @@ static struct rpmsg_device *rpmsg_virtio_add_ctrl_dev(struct virtio_device *vdev
 
 	err = rpmsg_chrdev_register_device(rpdev_ctrl);
 	if (err) {
-		kfree(vch);
+		/* vch will be free in virtio_rpmsg_release_device() */
 		return ERR_PTR(err);
 	}
 
Signature · 1 line
2.35.1
author maintainer
#29 lore raw

PATCH 29/298 rpmsg: virtio: Fix the unregistration of the device rpmsg_ctrl

In reply to Greg Kroah-Hartman

From: Arnaud Pouliquen <[email protected]>

[ Upstream commit df191796985922488e4e6b64f7bd79c3934412f2 ]

Unregister the rpmsg_ctrl device instead of just freeing the the virtio_rpmsg_channel structure.
This will properly unregister the device and call virtio_rpmsg_release_device() that frees the structure.

Fixes c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
---
 drivers/rpmsg/virtio_rpmsg_bus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rpmsg/virtio_rpmsg_bus.c +1 -1
index 99e0de65733a..ed26e9226834 100644
@@ -862,7 +862,7 @@ static void rpmsg_virtio_del_ctrl_dev(struct rpmsg_device *rpdev_ctrl)
 {
 	if (!rpdev_ctrl)
 		return;
-	kfree(to_virtio_rpmsg_channel(rpdev_ctrl));
+	device_unregister(&rpdev_ctrl->dev);
 }
 
 static int rpmsg_probe(struct virtio_device *vdev)
Signature · 1 line
2.35.1
author maintainer
#30 lore raw

PATCH 30/298 iio: adc: stmpe-adc: Fix wait_for_completion_timeout return value check

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit d345b23200bcdbd2bd3582213d738c258b77718f ]

wait_for_completion_timeout() returns unsigned long not long. it returns 0 if timed out, and positive if completed. The check for <= 0 is ambiguous and should be == 0 here indicating timeout which is the only error case

Fixes e813dde6f833 ("iio: stmpe-adc: Use wait_for_completion_timeout")
---
 drivers/iio/adc/stmpe-adc.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/iio/adc/stmpe-adc.c +4 -4
index d2d405388499..83e0ac4467ca 100644
@@ -61,7 +61,7 @@ struct stmpe_adc {
 static int stmpe_read_voltage(struct stmpe_adc *info,
 		struct iio_chan_spec const *chan, int *val)
 {
-	long ret;
+	unsigned long ret;
 
 	mutex_lock(&info->lock);
 
@@ -79,7 +79,7 @@ static int stmpe_read_voltage(struct stmpe_adc *info,
 
 	ret = wait_for_completion_timeout(&info->completion, STMPE_ADC_TIMEOUT);
 
-	if (ret <= 0) {
+	if (ret == 0) {
 		stmpe_reg_write(info->stmpe, STMPE_REG_ADC_INT_STA,
 				STMPE_ADC_CH(info->channel));
 		mutex_unlock(&info->lock);
@@ -96,7 +96,7 @@ static int stmpe_read_voltage(struct stmpe_adc *info,
 static int stmpe_read_temp(struct stmpe_adc *info,
 		struct iio_chan_spec const *chan, int *val)
 {
-	long ret;
+	unsigned long ret;
 
 	mutex_lock(&info->lock);
 
@@ -114,7 +114,7 @@ static int stmpe_read_temp(struct stmpe_adc *info,
 
 	ret = wait_for_completion_timeout(&info->completion, STMPE_ADC_TIMEOUT);
 
-	if (ret <= 0) {
+	if (ret == 0) {
 		mutex_unlock(&info->lock);
 		return -ETIMEDOUT;
 	}
Signature · 1 line
2.35.1
author maintainer
#31 lore raw

PATCH 31/298 iio: proximity: vl53l0x: Fix return value check of wait_for_completion_timeout

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 50f2959113cb6756ffd73c4fedc712cf2661f711 ]

wait_for_completion_timeout() returns unsigned long not int. It returns 0 if timed out, and positive if completed. The check for <= 0 is ambiguous and should be == 0 here indicating timeout which is the only error case.

Fixes 3cef2e31b54b ("iio: proximity: vl53l0x: Add IRQ support")
---
 drivers/iio/proximity/vl53l0x-i2c.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/iio/proximity/vl53l0x-i2c.c +3 -4
index cf38144b6f95..13a87d3e3544 100644
@@ -104,6 +104,7 @@ static int vl53l0x_read_proximity(struct vl53l0x_data *data,
 	u16 tries = 20;
 	u8 buffer[12];
 	int ret;
+	unsigned long time_left;
 
 	ret = i2c_smbus_write_byte_data(client, VL_REG_SYSRANGE_START, 1);
 	if (ret < 0)
@@ -112,10 +113,8 @@ static int vl53l0x_read_proximity(struct vl53l0x_data *data,
 	if (data->client->irq) {
 		reinit_completion(&data->completion);
 
-		ret = wait_for_completion_timeout(&data->completion, HZ/10);
-		if (ret < 0)
-			return ret;
-		else if (ret == 0)
+		time_left = wait_for_completion_timeout(&data->completion, HZ/10);
+		if (time_left == 0)
 			return -ETIMEDOUT;
 
 		vl53l0x_clear_irq(data);
Signature · 1 line
2.35.1
author maintainer
#32 lore raw

PATCH 32/298 iio: adc: sc27xx: fix read big scale voltage not right

In reply to Greg Kroah-Hartman

From: Cixi Geng <[email protected]>

[ Upstream commit ad930a75613282400179361e220e58b87386b8c7 ]

Fix wrong configuration value of SC27XX_ADC_SCALE_MASK and SC27XX_ADC_SCALE_SHIFT by spec documetation.

Fixes 5df362a6cf49c (iio: adc: Add Spreadtrum SC27XX PMICs ADC support)
---
 drivers/iio/adc/sc27xx_adc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/iio/adc/sc27xx_adc.c +2 -2
index 00098caf6d9e..aee076c8e2b1 100644
@@ -36,8 +36,8 @@
 
 /* Bits and mask definition for SC27XX_ADC_CH_CFG register */
 #define SC27XX_ADC_CHN_ID_MASK		GENMASK(4, 0)
-#define SC27XX_ADC_SCALE_MASK		GENMASK(10, 8)
-#define SC27XX_ADC_SCALE_SHIFT		8
+#define SC27XX_ADC_SCALE_MASK		GENMASK(10, 9)
+#define SC27XX_ADC_SCALE_SHIFT		9
 
 /* Bits definitions for SC27XX_ADC_INT_EN registers */
 #define SC27XX_ADC_IRQ_EN		BIT(0)
Signature · 1 line
2.35.1
author maintainer
#33 lore raw

PATCH 33/298 iio: adc: sc27xx: Fine tune the scale calibration values

In reply to Greg Kroah-Hartman

From: Cixi Geng <[email protected]>

[ Upstream commit 5a7a184b11c6910f47600ff5cbbee34168f701a8 ]

Small adjustment the scale calibration value for the sc2731, use new name sc2731_[big|small]_scale_graph_calib, and remove the origin [big|small]_scale_graph_calib struct for unused.

Fixes 8ba0dbfd07a35 (iio: adc: sc27xx: Add ADC scale calibration)
---
 drivers/iio/adc/sc27xx_adc.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/iio/adc/sc27xx_adc.c +8 -8
index aee076c8e2b1..cfe003cc4f0b 100644
@@ -103,14 +103,14 @@ static struct sc27xx_adc_linear_graph small_scale_graph = {
 	100, 341,
 };
 
-static const struct sc27xx_adc_linear_graph big_scale_graph_calib = {
-	4200, 856,
-	3600, 733,
+static const struct sc27xx_adc_linear_graph sc2731_big_scale_graph_calib = {
+	4200, 850,
+	3600, 728,
 };
 
-static const struct sc27xx_adc_linear_graph small_scale_graph_calib = {
-	1000, 833,
-	100, 80,
+static const struct sc27xx_adc_linear_graph sc2731_small_scale_graph_calib = {
+	1000, 838,
+	100, 84,
 };
 
 static int sc27xx_adc_get_calib_data(u32 calib_data, int calib_adc)
@@ -130,11 +130,11 @@ static int sc27xx_adc_scale_calibration(struct sc27xx_adc_data *data,
 	size_t len;
 
 	if (big_scale) {
-		calib_graph = &big_scale_graph_calib;
+		calib_graph = &sc2731_big_scale_graph_calib;
 		graph = &big_scale_graph;
 		cell_name = "big_scale_calib";
 	} else {
-		calib_graph = &small_scale_graph_calib;
+		calib_graph = &sc2731_small_scale_graph_calib;
 		graph = &small_scale_graph;
 		cell_name = "small_scale_calib";
 	}
Signature · 1 line
2.35.1
author maintainer
#34 lore raw

PATCH 34/298 rpmsg: qcom_smd: Fix returning 0 if irq_of_parse_and_map() fails

In reply to Greg Kroah-Hartman

From: Krzysztof Kozlowski <[email protected]>

[ Upstream commit 59d6f72f6f9c92fec8757d9e29527da828e9281f ]

irq_of_parse_and_map() returns 0 on failure, so this should not be passed further as error return code.

Fixes 1a358d350664 ("rpmsg: qcom_smd: Fix irq_of_parse_and_map() return value")
---
 drivers/rpmsg/qcom_smd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rpmsg/qcom_smd.c +1 -1
index d4b54eebe15d..4ad90945518f 100644
@@ -1406,7 +1406,7 @@ static int qcom_smd_parse_edge(struct device *dev,
 	irq = irq_of_parse_and_map(node, 0);
 	if (!irq) {
 		dev_err(dev, "required smd interrupt missing\n");
-		ret = irq;
+		ret = -EINVAL;
 		goto put_node;
 	}
 
Signature · 1 line
2.35.1
author maintainer
#35 lore raw

PATCH 35/298 misc/pvpanic: Convert regular spinlock into trylock on panic path

In reply to Greg Kroah-Hartman

From: Guilherme G. Piccoli <[email protected]>

[ Upstream commit e918c10265ef2bc82ce8a6fed6d8123d09ec1db3 ]

The pvpanic driver relies on panic notifiers to execute a callback on panic event. Such function is executed in atomic context - the panic function disables local IRQs, preemption and all other CPUs that aren't running the panic code.

With that said, it's dangerous to use regular spinlocks in such path, as introduced by commit b3c0f8774668 ("misc/pvpanic: probe multiple instances"). This patch fixes that by replacing regular spinlocks with the trylock safer approach.

It also fixes an old comment (about a long gone framebuffer code) and the notifier priority - we should execute hypervisor notifiers early, deferring this way the panic action to the hypervisor, as expected by the users that are setting up pvpanic.

Fixes b3c0f8774668 ("misc/pvpanic: probe multiple instances")
Cc Christophe JAILLET <[email protected]>
Cc Mihai Carabas <[email protected]>
Cc Shile Zhang <[email protected]>
Cc Wang ShaoBo <[email protected]>
Cc zhenwei pi <[email protected]>
---
 drivers/misc/pvpanic/pvpanic.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/misc/pvpanic/pvpanic.c +8 -2
index 4b8f1c7d726d..049a12006348 100644
@@ -34,7 +34,9 @@ pvpanic_send_event(unsigned int event)
 {
 	struct pvpanic_instance *pi_cur;
 
-	spin_lock(&pvpanic_lock);
+	if (!spin_trylock(&pvpanic_lock))
+		return;
+
 	list_for_each_entry(pi_cur, &pvpanic_list, list) {
 		if (event & pi_cur->capability & pi_cur->events)
 			iowrite8(event, pi_cur->base);
@@ -55,9 +57,13 @@ pvpanic_panic_notify(struct notifier_block *nb, unsigned long code, void *unused
 	return NOTIFY_DONE;
 }
 
+/*
+ * Call our notifier very early on panic, deferring the
+ * action taken to the hypervisor.
+ */
 static struct notifier_block pvpanic_panic_nb = {
 	.notifier_call = pvpanic_panic_notify,
-	.priority = 1, /* let this called before broken drm_fb_helper() */
+	.priority = INT_MAX,
 };
 
 static void pvpanic_remove(void *param)
Signature · 1 line
2.35.1
author maintainer
#36 lore raw

PATCH 36/298 phy: qcom-qmp: fix pipe-clock imbalance on power-on failure

In reply to Greg Kroah-Hartman

From: Johan Hovold <[email protected]>

[ Upstream commit 5e73b2d9867998278479ccc065a8a8227a5513ef ]

Make sure to disable the pipe clock also if ufs-reset deassertion fails during power on.

Note that the ufs-reset is asserted in qcom_qmp_phy_com_exit().

Fixes c9b589791fc1 ("phy: qcom: Utilize UFS reset controller")
Cc Evan Green <[email protected]>
---
 drivers/phy/qualcomm/phy-qcom-qmp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/qualcomm/phy-qcom-qmp.c +1 -1
index c2b878128e2c..7493fd634c1d 100644
@@ -5246,7 +5246,7 @@ static int qcom_qmp_phy_power_on(struct phy *phy)
 
 	ret = reset_control_deassert(qmp->ufs_reset);
 	if (ret)
-		goto err_lane_rst;
+		goto err_pcs_ready;
 
 	qcom_qmp_phy_configure(pcs_misc, cfg->regs, cfg->pcs_misc_tbl,
 			       cfg->pcs_misc_tbl_num);
Signature · 1 line
2.35.1
author maintainer
#37 lore raw

PATCH 37/298 power: supply: axp288_fuel_gauge: Drop BIOS version check from "T3 MRD" DMI quirk

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit f61509a6f0b70f5bedea34efaf8065621689bd7a ]

Some "T3 MRD" mini-PCs / HDMI-sticks without a battery use a different value then "5.11" for their DMI BIOS version field.

Drop the BIOS version check so that the no-battery "T3 MRD" DMI quirk applies to these too.

Fixes 3a06b912a5ce ("power: supply: axp288_fuel_gauge: Make "T3 MRD" no_battery_list DMI entry more generic")
---
 drivers/power/supply/axp288_fuel_gauge.c | 1 -
 1 file changed, 1 deletion(-)
drivers/power/supply/axp288_fuel_gauge.c +0 -1
index ce8ffd0a41b5..68595897e72d 100644
@@ -600,7 +600,6 @@ static const struct dmi_system_id axp288_no_battery_list[] = {
 			DMI_MATCH(DMI_BOARD_NAME, "T3 MRD"),
 			DMI_MATCH(DMI_CHASSIS_TYPE, "3"),
 			DMI_MATCH(DMI_BIOS_VENDOR, "American Megatrends Inc."),
-			DMI_MATCH(DMI_BIOS_VERSION, "5.11"),
 		},
 	},
 	{}
Signature · 1 line
2.35.1
author maintainer
#38 lore raw

PATCH 38/298 power: supply: ab8500_fg: Allocate wq in probe

In reply to Greg Kroah-Hartman

From: Linus Walleij <[email protected]>

[ Upstream commit 010ddb813f3554cbbf8bd13b731452236a2c8017 ]

The workqueue is allocated in bind() but all interrupts are registered in probe().

Some interrupts put work on the workqueue, which can have bad side effects.

Allocate the workqueue in probe() instead, destroy it in .remove() and make unbind() simply flush the workqueue.

Fixes 1c1f13a006ed ("power: supply: ab8500: Move to componentized binding")
---
 drivers/power/supply/ab8500_fg.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)
drivers/power/supply/ab8500_fg.c +10 -9
index 09a4cbd69676..23adcb597ff9 100644
@@ -2995,13 +2995,6 @@ static int ab8500_fg_bind(struct device *dev, struct device *master,
 {
 	struct ab8500_fg *di = dev_get_drvdata(dev);
 
-	/* Create a work queue for running the FG algorithm */
-	di->fg_wq = alloc_ordered_workqueue("ab8500_fg_wq", WQ_MEM_RECLAIM);
-	if (di->fg_wq == NULL) {
-		dev_err(dev, "failed to create work queue\n");
-		return -ENOMEM;
-	}
-
 	di->bat_cap.max_mah_design = di->bm->bi->charge_full_design_uah;
 	di->bat_cap.max_mah = di->bat_cap.max_mah_design;
 	di->vbat_nom_uv = di->bm->bi->voltage_max_design_uv;
@@ -3025,8 +3018,7 @@ static void ab8500_fg_unbind(struct device *dev, struct device *master,
 	if (ret)
 		dev_err(dev, "failed to disable coulomb counter\n");
 
-	destroy_workqueue(di->fg_wq);
-	flush_scheduled_work();
+	flush_workqueue(di->fg_wq);
 }
 
 static const struct component_ops ab8500_fg_component_ops = {
@@ -3070,6 +3062,13 @@ static int ab8500_fg_probe(struct platform_device *pdev)
 	ab8500_fg_charge_state_to(di, AB8500_FG_CHARGE_INIT);
 	ab8500_fg_discharge_state_to(di, AB8500_FG_DISCHARGE_INIT);
 
+	/* Create a work queue for running the FG algorithm */
+	di->fg_wq = alloc_ordered_workqueue("ab8500_fg_wq", WQ_MEM_RECLAIM);
+	if (di->fg_wq == NULL) {
+		dev_err(dev, "failed to create work queue\n");
+		return -ENOMEM;
+	}
+
 	/* Init work for running the fg algorithm instantly */
 	INIT_WORK(&di->fg_work, ab8500_fg_instant_work);
 
@@ -3181,6 +3180,8 @@ static int ab8500_fg_remove(struct platform_device *pdev)
 	int ret = 0;
 	struct ab8500_fg *di = platform_get_drvdata(pdev);
 
+	destroy_workqueue(di->fg_wq);
+	flush_scheduled_work();
 	component_del(&pdev->dev, &ab8500_fg_component_ops);
 	list_del(&di->node);
 	ab8500_fg_sysfs_exit(di);
Signature · 1 line
2.35.1
author maintainer
#39 lore raw

PATCH 39/298 serial: sifive: Report actual baud base rather than fixed 115200

In reply to Greg Kroah-Hartman

From: Maciej W. Rozycki <[email protected]>

[ Upstream commit 0a7ff843d507ce2cca2c3b7e169ee56e28133530 ]

The base baud value reported is supposed to be the highest baud rate that can be set for a serial port. The SiFive FU740-C000 SOC's on-chip UART supports baud rates of up to 1/16 of the input clock rate, which is the bus clock `tlclk'[1], often at 130MHz in the case of the HiFive Unmatched board.

However the sifive UART driver reports a fixed value of 115200 instead:

10010000.serial: ttySIF0 at MMIO 0x10010000 (irq = 1, base_baud = 115200) is a SiFive UART v0 10011000.serial: ttySIF1 at MMIO 0x10011000 (irq = 2, base_baud = 115200) is a SiFive UART v0

even though we already support setting higher baud rates, e.g.:

$ tty
/dev/ttySIF1
$ stty speed
230400

The baud base value is computed by the serial core by dividing the UART clock recorded in `struct uart_port' by 16, which is also the minimum value of the clock divider supported, so correct the baud base value reported by setting the UART clock recorded to the input clock rate rather than 115200:

10010000.serial: ttySIF0 at MMIO 0x10010000 (irq = 1, base_baud = 8125000) is a SiFive UART v0 10011000.serial: ttySIF1 at MMIO 0x10011000 (irq = 2, base_baud = 8125000) is a SiFive UART v0

[1] "SiFive FU740-C000 Manual", v1p3, SiFive, Inc., August 13, 2021,

    Section 16.9 "Baud Rate Divisor Register (div)", pp.143-144
---
 drivers/tty/serial/sifive.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/tty/serial/sifive.c +1 -1
index b79900d0e91a..fff0b7916f3d 100644
@@ -998,7 +998,7 @@ static int sifive_serial_probe(struct platform_device *pdev)
 	/* Set up clock divider */
 	ssp->clkin_rate = clk_get_rate(ssp->clk);
 	ssp->baud_rate = SIFIVE_DEFAULT_BAUD_RATE;
-	ssp->port.uartclk = ssp->baud_rate * 16;
+	ssp->port.uartclk = ssp->clkin_rate;
 	__ssp_update_div(ssp);
 
 	platform_set_drvdata(pdev, ssp);
Signature · 1 line
2.35.1
author maintainer
#40 lore raw

PATCH 40/298 export: fix string handling of namespace in EXPORT_SYMBOL_NS

In reply to Greg Kroah-Hartman

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

[ Upstream commit d143b9db8069f0e2a0fa34484e806a55a0dd4855 ]

Commit c3a6cf19e695 ("export: avoid code duplication in include/linux/export.h") broke the ability for a defined string to be used as a namespace value. Fix this up by using stringify to properly encode the namespace name.

Fixes c3a6cf19e695 ("export: avoid code duplication in include/linux/export.h")
Cc Miroslav Benes <[email protected]>
Cc Emil Velikov <[email protected]>
Cc Jessica Yu <[email protected]>
Cc Quentin Perret <[email protected]>
Cc Matthias Maennich <[email protected]>
---
 include/linux/export.h | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
include/linux/export.h +4 -3
index 27d848712b90..5910ccb66ca2 100644
@@ -2,6 +2,8 @@
 #ifndef _LINUX_EXPORT_H
 #define _LINUX_EXPORT_H
 
+#include <linux/stringify.h>
+
 /*
  * Export symbols from the kernel to modules.  Forked from module.h
  * to reduce the amount of pointless cruft we feed to gcc when only
@@ -154,7 +156,6 @@ struct kernel_symbol {
 #endif /* CONFIG_MODULES */
 
 #ifdef DEFAULT_SYMBOL_NAMESPACE
-#include <linux/stringify.h>
 #define _EXPORT_SYMBOL(sym, sec)	__EXPORT_SYMBOL(sym, sec, __stringify(DEFAULT_SYMBOL_NAMESPACE))
 #else
 #define _EXPORT_SYMBOL(sym, sec)	__EXPORT_SYMBOL(sym, sec, "")
@@ -162,8 +163,8 @@ struct kernel_symbol {
 
 #define EXPORT_SYMBOL(sym)		_EXPORT_SYMBOL(sym, "")
 #define EXPORT_SYMBOL_GPL(sym)		_EXPORT_SYMBOL(sym, "_gpl")
-#define EXPORT_SYMBOL_NS(sym, ns)	__EXPORT_SYMBOL(sym, "", #ns)
-#define EXPORT_SYMBOL_NS_GPL(sym, ns)	__EXPORT_SYMBOL(sym, "_gpl", #ns)
+#define EXPORT_SYMBOL_NS(sym, ns)	__EXPORT_SYMBOL(sym, "", __stringify(ns))
+#define EXPORT_SYMBOL_NS_GPL(sym, ns)	__EXPORT_SYMBOL(sym, "_gpl", __stringify(ns))
 
 #endif /* !__ASSEMBLY__ */
 
Signature · 1 line
2.35.1
author maintainer
#41 lore raw

PATCH 41/298 watchdog: rzg2l_wdt: Fix 32bit overflow issue

In reply to Greg Kroah-Hartman

From: Biju Das <[email protected]>

[ Upstream commit ea2949df22a533cdf75e4583c00b1ce94cd5a83b ]

The value of timer_cycle_us can be 0 due to 32bit overflow. For eg:- If we assign the counter value "0xfff" for computing maxval.

This patch fixes this issue by appending ULL to 1024, so that it is promoted to 64bit.

This patch also fixes the warning message, 'watchdog: Invalid min and max timeout values, resetting to 0!'.

---
 drivers/watchdog/rzg2l_wdt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/watchdog/rzg2l_wdt.c +1 -1
index 6b426df34fd6..96f2a018ab62 100644
@@ -53,7 +53,7 @@ static void rzg2l_wdt_wait_delay(struct rzg2l_wdt_priv *priv)
 
 static u32 rzg2l_wdt_get_cycle_usec(unsigned long cycle, u32 wdttime)
 {
-	u64 timer_cycle_us = 1024 * 1024 * (wdttime + 1) * MICRO;
+	u64 timer_cycle_us = 1024 * 1024ULL * (wdttime + 1) * MICRO;
 
 	return div64_ul(timer_cycle_us, cycle);
 }
Signature · 1 line
2.35.1
author maintainer
#42 lore raw

PATCH 42/298 watchdog: rzg2l_wdt: Fix Runtime PM usage

In reply to Greg Kroah-Hartman

From: Biju Das <[email protected]>

[ Upstream commit 95abafe76297fa057de6c3486ef844bd446bdf18 ]

Both rzg2l_wdt_probe() and rzg2l_wdt_start() calls pm_runtime_get() which results in a usage counter imbalance. This patch fixes this issue by removing pm_runtime_get() call from probe.

---
 drivers/watchdog/rzg2l_wdt.c | 16 ++--------------
 1 file changed, 2 insertions(+), 14 deletions(-)
drivers/watchdog/rzg2l_wdt.c +2 -14
index 96f2a018ab62..0fc73b8a9567 100644
@@ -151,12 +151,11 @@ static const struct watchdog_ops rzg2l_wdt_ops = {
 	.restart = rzg2l_wdt_restart,
 };
 
-static void rzg2l_wdt_reset_assert_pm_disable_put(void *data)
+static void rzg2l_wdt_reset_assert_pm_disable(void *data)
 {
 	struct watchdog_device *wdev = data;
 	struct rzg2l_wdt_priv *priv = watchdog_get_drvdata(wdev);
 
-	pm_runtime_put(wdev->parent);
 	pm_runtime_disable(wdev->parent);
 	reset_control_assert(priv->rstc);
 }
@@ -206,11 +205,6 @@ static int rzg2l_wdt_probe(struct platform_device *pdev)
 
 	reset_control_deassert(priv->rstc);
 	pm_runtime_enable(&pdev->dev);
-	ret = pm_runtime_resume_and_get(&pdev->dev);
-	if (ret < 0) {
-		dev_err(dev, "pm_runtime_resume_and_get failed ret=%pe", ERR_PTR(ret));
-		goto out_pm_get;
-	}
 
 	priv->wdev.info = &rzg2l_wdt_ident;
 	priv->wdev.ops = &rzg2l_wdt_ops;
@@ -222,7 +216,7 @@ static int rzg2l_wdt_probe(struct platform_device *pdev)
 
 	watchdog_set_drvdata(&priv->wdev, priv);
 	ret = devm_add_action_or_reset(&pdev->dev,
-				       rzg2l_wdt_reset_assert_pm_disable_put,
+				       rzg2l_wdt_reset_assert_pm_disable,
 				       &priv->wdev);
 	if (ret < 0)
 		return ret;
@@ -235,12 +229,6 @@ static int rzg2l_wdt_probe(struct platform_device *pdev)
 		dev_warn(dev, "Specified timeout invalid, using default");
 
 	return devm_watchdog_register_device(&pdev->dev, &priv->wdev);
-
-out_pm_get:
-	pm_runtime_disable(dev);
-	reset_control_assert(priv->rstc);
-
-	return ret;
 }
 
 static const struct of_device_id rzg2l_wdt_ids[] = {
Signature · 1 line
2.35.1
author maintainer
#43 lore raw

PATCH 43/298 watchdog: rzg2l_wdt: Fix BUG: Invalid wait context

In reply to Greg Kroah-Hartman

From: Biju Das <[email protected]>

[ Upstream commit e4cf89596c1f1e33309556699f910ced4abbaf44 ]

This patch fixes the issue 'BUG: Invalid wait context' during restart() callback by using clk_prepare_enable() instead of pm_runtime_get_sync() for turning on the clocks during restart.

This issue is noticed when testing with renesas_defconfig.

[   42.213802] reboot: Restarting system
[   42.217860]
[   42.219364] =============================
[   42.223368] [ BUG: Invalid wait context ]
[   42.227372] 5.17.0-rc5-arm64-renesas-00002-g10393723e35e #522 Not tainted
[   42.234153] -----------------------------
[   42.238155] systemd-shutdow/1 is trying to lock:
[   42.242766] ffff00000a650828 (&genpd->mlock){+.+.}-{3:3}, at: genpd_lock_mtx+0x14/0x20
[   42.250709] other info that might help us debug this:
[   42.255753] context-{4:4}
[   42.258368] 2 locks held by systemd-shutdow/1:
[   42.262806]  #0: ffff80000944e1c8 (system_transition_mutex#2){+.+.}-{3:3}, at: __do_sys_reboot+0xd0/0x250
[   42.272388]  #1: ffff8000094c4e40 (rcu_read_lock){....}-{1:2}, at: atomic_notifier_call_chain+0x0/0x150
[   42.281795] stack backtrace:
[   42.284672] CPU: 0 PID: 1 Comm: systemd-shutdow Not tainted 5.17.0-rc5-arm64-renesas-00002-g10393723e35e #522
[   42.294577] Hardware name: Renesas SMARC EVK based on r9a07g044c2 (DT)
[   42.301096] Call trace:
[   42.303538]  dump_backtrace+0xcc/0xd8
[   42.307203]  show_stack+0x14/0x30
[   42.310517]  dump_stack_lvl+0x88/0xb0
[   42.314180]  dump_stack+0x14/0x2c
[   42.317492]  __lock_acquire+0x1b24/0x1b50
[   42.321502]  lock_acquire+0x120/0x3a8
[   42.325162]  __mutex_lock+0x84/0x8f8
[   42.328737]  mutex_lock_nested+0x30/0x58
[   42.332658]  genpd_lock_mtx+0x14/0x20
[   42.336319]  genpd_runtime_resume+0xc4/0x228
[   42.340587]  __rpm_callback+0x44/0x170
[   42.344337]  rpm_callback+0x64/0x70
[   42.347824]  rpm_resume+0x4e0/0x6b8
[   42.351310]  __pm_runtime_resume+0x50/0x78
[   42.355404]  rzg2l_wdt_restart+0x28/0x68
[   42.359329]  watchdog_restart_notifier+0x1c/0x30
[   42.363943]  atomic_notifier_call_chain+0x94/0x150
[   42.368732]  do_kernel_restart+0x24/0x30
[   42.372652]  machine_restart+0x44/0x70
[   42.376399]  kernel_restart+0x3c/0x60
[   42.380058]  __do_sys_reboot+0x228/0x250
[   42.383977]  __arm64_sys_reboot+0x20/0x28
[   42.387983]  invoke_syscall+0x40/0xf8
---
 drivers/watchdog/rzg2l_wdt.c | 25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)
drivers/watchdog/rzg2l_wdt.c +13 -12
index 0fc73b8a9567..48dfe6e5e64f 100644
@@ -43,6 +43,8 @@ struct rzg2l_wdt_priv {
 	struct reset_control *rstc;
 	unsigned long osc_clk_rate;
 	unsigned long delay;
+	struct clk *pclk;
+	struct clk *osc_clk;
 };
 
 static void rzg2l_wdt_wait_delay(struct rzg2l_wdt_priv *priv)
@@ -118,7 +120,9 @@ static int rzg2l_wdt_restart(struct watchdog_device *wdev,
 
 	/* Reset the module before we modify any register */
 	reset_control_reset(priv->rstc);
-	pm_runtime_get_sync(wdev->parent);
+
+	clk_prepare_enable(priv->pclk);
+	clk_prepare_enable(priv->osc_clk);
 
 	/* smallest counter value to reboot soon */
 	rzg2l_wdt_write(priv, WDTSET_COUNTER_VAL(1), WDTSET);
@@ -165,7 +169,6 @@ static int rzg2l_wdt_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct rzg2l_wdt_priv *priv;
 	unsigned long pclk_rate;
-	struct clk *wdt_clk;
 	int ret;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
@@ -177,22 +180,20 @@ static int rzg2l_wdt_probe(struct platform_device *pdev)
 		return PTR_ERR(priv->base);
 
 	/* Get watchdog main clock */
-	wdt_clk = clk_get(&pdev->dev, "oscclk");
-	if (IS_ERR(wdt_clk))
-		return dev_err_probe(&pdev->dev, PTR_ERR(wdt_clk), "no oscclk");
+	priv->osc_clk = devm_clk_get(&pdev->dev, "oscclk");
+	if (IS_ERR(priv->osc_clk))
+		return dev_err_probe(&pdev->dev, PTR_ERR(priv->osc_clk), "no oscclk");
 
-	priv->osc_clk_rate = clk_get_rate(wdt_clk);
-	clk_put(wdt_clk);
+	priv->osc_clk_rate = clk_get_rate(priv->osc_clk);
 	if (!priv->osc_clk_rate)
 		return dev_err_probe(&pdev->dev, -EINVAL, "oscclk rate is 0");
 
 	/* Get Peripheral clock */
-	wdt_clk = clk_get(&pdev->dev, "pclk");
-	if (IS_ERR(wdt_clk))
-		return dev_err_probe(&pdev->dev, PTR_ERR(wdt_clk), "no pclk");
+	priv->pclk = devm_clk_get(&pdev->dev, "pclk");
+	if (IS_ERR(priv->pclk))
+		return dev_err_probe(&pdev->dev, PTR_ERR(priv->pclk), "no pclk");
 
-	pclk_rate = clk_get_rate(wdt_clk);
-	clk_put(wdt_clk);
+	pclk_rate = clk_get_rate(priv->pclk);
 	if (!pclk_rate)
 		return dev_err_probe(&pdev->dev, -EINVAL, "pclk rate is 0");
 
Signature · 1 line
2.35.1
author maintainer
#44 lore raw

PATCH 44/298 watchdog: rzg2l_wdt: Fix reset control imbalance

In reply to Greg Kroah-Hartman

From: Biju Das <[email protected]>

[ Upstream commit 33d04d0fdba9fae18c7d58364643d2c606a43dba ]

Both rzg2l_wdt_probe() and rzg2l_wdt_start() calls reset_control_ deassert() which results in a reset control imbalance.

This patch fixes reset control imbalance by removing reset_control_ deassert() from rzg2l_wdt_start() and replaces reset_control_assert with reset_control_reset in rzg2l_wdt_stop() as watchdog module can be stopped only by a module reset. This change will allow us to restart WDT after stop() by configuring WDT timeout and enable registers.

---
 drivers/watchdog/rzg2l_wdt.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/watchdog/rzg2l_wdt.c +1 -2
index 48dfe6e5e64f..88274704b260 100644
@@ -88,7 +88,6 @@ static int rzg2l_wdt_start(struct watchdog_device *wdev)
 {
 	struct rzg2l_wdt_priv *priv = watchdog_get_drvdata(wdev);
 
-	reset_control_deassert(priv->rstc);
 	pm_runtime_get_sync(wdev->parent);
 
 	/* Initialize time out */
@@ -108,7 +107,7 @@ static int rzg2l_wdt_stop(struct watchdog_device *wdev)
 	struct rzg2l_wdt_priv *priv = watchdog_get_drvdata(wdev);
 
 	pm_runtime_put(wdev->parent);
-	reset_control_assert(priv->rstc);
+	reset_control_reset(priv->rstc);
 
 	return 0;
 }
Signature · 1 line
2.35.1
author maintainer
#45 lore raw

PATCH 45/298 soundwire: intel: prevent pm_runtime resume prior to system suspend

In reply to Greg Kroah-Hartman

From: Pierre-Louis Bossart <[email protected]>

[ Upstream commit 6d9f2dadba698114fed97b224578c5338a36b0d9 ]

commit e38f9ff63e6d ("ACPI: scan: Do not add device IDs from _CID if _HID is not valid") exposes a race condition on a TGL RVP device leading to a timeout.

The detailed analysis shows the RT711 codec driver scheduling a jack detection workqueue while attaching during a spurious pm_runtime resume, and the work function happens to be scheduled after the manager device is suspended.

The direct link between this ACPI patch and a spurious pm_runtime resume is not obvious; the most likely explanation is that a change in the ACPI device linked list management modifies the order in which the pm_runtime device status is checked and exposes a race condition that was probably present for a very long time, but was not identified.

We already have a check in the .prepare stage, where we will resume to full power from specific clock-stop modes. In all other cases, we don't need to resume to full power by default. Adding the SMART_SUSPEND flag prevents the spurious resume from happening.

---
 drivers/soundwire/intel.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/soundwire/intel.c +3 -0
index 63101f1ba271..32e5fdb823c4 100644
@@ -1293,6 +1293,9 @@ static int intel_link_probe(struct auxiliary_device *auxdev,
 	/* use generic bandwidth allocation algorithm */
 	sdw->cdns.bus.compute_params = sdw_compute_params;
 
+	/* avoid resuming from pm_runtime suspend if it's not required */
+	dev_pm_set_driver_flags(dev, DPM_FLAG_SMART_SUSPEND);
+
 	ret = sdw_bus_master_add(bus, dev, dev->fwnode);
 	if (ret) {
 		dev_err(dev, "sdw_bus_master_add fail: %d\n", ret);
Signature · 1 line
2.35.1
author maintainer
#46 lore raw

PATCH 46/298 coresight: cpu-debug: Replace mutex with mutex_trylock on panic notifier

In reply to Greg Kroah-Hartman

From: Guilherme G. Piccoli <[email protected]>

[ Upstream commit 1adff542d67a2ed1120955cb219bfff8a9c53f59 ]

The panic notifier infrastructure executes registered callbacks when a panic event happens - such callbacks are executed in atomic context, with interrupts and preemption disabled in the running CPU and all other CPUs disabled. That said, mutexes in such context are not a good idea.

This patch replaces a regular mutex with a mutex_trylock safer approach; given the nature of the mutex used in the driver, it should be pretty uncommon being unable to acquire such mutex in the panic path, hence no functional change should be observed (and if it is, that would be likely a deadlock with the regular mutex).

Fixes 2227b7c74634 ("coresight: add support for CPU debug module")
Cc Leo Yan <[email protected]>
Cc Mathieu Poirier <[email protected]>
Cc Mike Leach <[email protected]>
Cc Suzuki K Poulose <[email protected]>
---
 drivers/hwtracing/coresight/coresight-cpu-debug.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/hwtracing/coresight/coresight-cpu-debug.c +4 -3
index 8845ec4b4402..1874df7c6a73 100644
@@ -380,9 +380,10 @@ static int debug_notifier_call(struct notifier_block *self,
 	int cpu;
 	struct debug_drvdata *drvdata;
 
-	mutex_lock(&debug_lock);
+	/* Bail out if we can't acquire the mutex or the functionality is off */
+	if (!mutex_trylock(&debug_lock))
+		return NOTIFY_DONE;
 
-	/* Bail out if the functionality is disabled */
 	if (!debug_enable)
 		goto skip_dump;
 
@@ -401,7 +402,7 @@ static int debug_notifier_call(struct notifier_block *self,
 
 skip_dump:
 	mutex_unlock(&debug_lock);
-	return 0;
+	return NOTIFY_DONE;
 }
 
 static struct notifier_block debug_notifier = {
Signature · 1 line
2.35.1
author maintainer
#47 lore raw

PATCH 47/298 ksmbd: fix reference count leak in smb_check_perm_dacl()

In reply to Greg Kroah-Hartman

From: Xin Xiong <[email protected]>

[ Upstream commit d21a580dafc69aa04f46e6099616146a536b0724 ]

The issue happens in a specific path in smb_check_perm_dacl(). When "id" and "uid" have the same value, the function simply jumps out of the loop without decrementing the reference count of the object "posix_acls", which is increased by get_acl() earlier. This may result in memory leaks.

Fix it by decreasing the reference count of "posix_acls" before jumping to label "check_access_bits".

Fixes 777cad1604d6 ("ksmbd: remove select FS_POSIX_ACL in Kconfig")
---
 fs/ksmbd/smbacl.c | 1 +
 1 file changed, 1 insertion(+)
fs/ksmbd/smbacl.c +1 -0
index 6ecf55ea1fed..38f23bf981ac 100644
@@ -1261,6 +1261,7 @@ int smb_check_perm_dacl(struct ksmbd_conn *conn, struct path *path,
 					if (!access_bits)
 						access_bits =
 							SET_MINIMUM_RIGHTS;
+					posix_acl_release(posix_acls);
 					goto check_access_bits;
 				}
 			}
Signature · 1 line
2.35.1
author maintainer
#48 lore raw

PATCH 48/298 extcon: ptn5150: Add queue work sync before driver release

In reply to Greg Kroah-Hartman

From: Li Jun <[email protected]>

[ Upstream commit 782cd939cbe0f569197cd1c9b0477ee213167f04 ]

Add device managed action to sync pending queue work, otherwise the queued work may run after the work is destroyed.

Fixes 4ed754de2d66 ("extcon: Add support for ptn5150 extcon driver")
---
 drivers/extcon/extcon-ptn5150.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/extcon/extcon-ptn5150.c +11 -0
index 5b9a3cf8df26..2a7874108df8 100644
@@ -194,6 +194,13 @@ static int ptn5150_init_dev_type(struct ptn5150_info *info)
 	return 0;
 }
 
+static void ptn5150_work_sync_and_put(void *data)
+{
+	struct ptn5150_info *info = data;
+
+	cancel_work_sync(&info->irq_work);
+}
+
 static int ptn5150_i2c_probe(struct i2c_client *i2c)
 {
 	struct device *dev = &i2c->dev;
@@ -284,6 +291,10 @@ static int ptn5150_i2c_probe(struct i2c_client *i2c)
 	if (ret)
 		return -EINVAL;
 
+	ret = devm_add_action_or_reset(dev, ptn5150_work_sync_and_put, info);
+	if (ret)
+		return ret;
+
 	/*
 	 * Update current extcon state if for example OTG connection was there
 	 * before the probe
Signature · 1 line
2.35.1
author maintainer
#49 lore raw

PATCH 49/298 dt-bindings: remoteproc: mediatek: Make l1tcm reg exclusive to mt819x

In reply to Greg Kroah-Hartman

From: Nícolas F. R. A. Prado <[email protected]>

[ Upstream commit 6bbe1065121b8cd3b3e734ef8cd99f142bdab241 ]

Commit ca23ecfdbd44 ("remoteproc/mediatek: support L1TCM") added support for the l1tcm memory region on the MT8192 SCP, adding a new da_to_va callback that handles l1tcm while keeping the old one for back-compatibility with MT8183. However, since the mt8192 compatible was missing from the dt-binding, the accompanying dt-binding commit 503c64cc42f1 ("dt-bindings: remoteproc: mediatek: add L1TCM memory region") mistakenly added this reg as if it were for mt8183. And later it became common to all platforms as their compatibles were added.

Fix the dt-binding so that the l1tcm reg can be present only on the supported platforms: mt8192 and mt8195.

Fixes 503c64cc42f1 ("dt-bindings: remoteproc: mediatek: add L1TCM memory region")
---
 .../bindings/remoteproc/mtk,scp.yaml          | 44 +++++++++++++------
 1 file changed, 30 insertions(+), 14 deletions(-)
Documentation/devicetree/bindings/remoteproc/mtk,scp.yaml +30 -14
index d21a25ee96e6..8b2c0f1f8550 100644
@@ -22,11 +22,13 @@ properties:
 
   reg:
     description:
-      Should contain the address ranges for memory regions SRAM, CFG, and
-      L1TCM.
+      Should contain the address ranges for memory regions SRAM, CFG, and,
+      on some platforms, L1TCM.
+    minItems: 2
     maxItems: 3
 
   reg-names:
+    minItems: 2
     items:
       - const: sram
       - const: cfg
@@ -46,16 +48,30 @@ required:
   - reg
   - reg-names
 
-if:
-  properties:
-    compatible:
-      enum:
-        - mediatek,mt8183-scp
-        - mediatek,mt8192-scp
-then:
-  required:
-    - clocks
-    - clock-names
+allOf:
+  - if:
+      properties:
+        compatible:
+          enum:
+            - mediatek,mt8183-scp
+            - mediatek,mt8192-scp
+    then:
+      required:
+        - clocks
+        - clock-names
+
+  - if:
+      properties:
+        compatible:
+          enum:
+            - mediatek,mt8183-scp
+            - mediatek,mt8186-scp
+    then:
+      properties:
+        reg:
+          maxItems: 2
+        reg-names:
+          maxItems: 2
 
 additionalProperties:
   type: object
@@ -75,10 +91,10 @@ additionalProperties:
 
 examples:
   - |
-    #include <dt-bindings/clock/mt8183-clk.h>
+    #include <dt-bindings/clock/mt8192-clk.h>
 
     scp@10500000 {
-        compatible = "mediatek,mt8183-scp";
+        compatible = "mediatek,mt8192-scp";
         reg = <0x10500000 0x80000>,
               <0x10700000 0x8000>,
               <0x10720000 0xe0000>;
Signature · 1 line
2.35.1
author maintainer
#50 lore raw

PATCH 50/298 soc: rockchip: Fix refcount leak in rockchip_grf_init

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 9b59588d8be91c96bfb0371e912ceb4f16315dbf ]

of_find_matching_node_and_match returns a node pointer with refcount incremented, we should use of_node_put() on it when done. Add missing of_node_put() to avoid refcount leak.

Fixes 4c58063d4258 ("soc: rockchip: add driver handling grf setup")
---
 drivers/soc/rockchip/grf.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/soc/rockchip/grf.c +2 -0
index 494cf2b5bf7b..343ff61ccccb 100644
@@ -148,12 +148,14 @@ static int __init rockchip_grf_init(void)
 		return -ENODEV;
 	if (!match || !match->data) {
 		pr_err("%s: missing grf data\n", __func__);
+		of_node_put(np);
 		return -EINVAL;
 	}
 
 	grf_info = match->data;
 
 	grf = syscon_node_to_regmap(np);
+	of_node_put(np);
 	if (IS_ERR(grf)) {
 		pr_err("%s: could not get grf syscon\n", __func__);
 		return PTR_ERR(grf);
Signature · 1 line
2.35.1
author maintainer
#51 lore raw

PATCH 51/298 clocksource/drivers/riscv: Events are stopped during CPU suspend

In reply to Greg Kroah-Hartman

From: Samuel Holland <[email protected]>

[ Upstream commit 232ccac1bd9b5bfe73895f527c08623e7fa0752d ]

Some implementations of the SBI time extension depend on hart-local state (for example, CSRs) that are lost or hardware that is powered down when a CPU is suspended. To be safe, the clockevents driver cannot assume that timer IRQs will be received during CPU suspend.

---
 drivers/clocksource/timer-riscv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clocksource/timer-riscv.c +1 -1
index 1767f8bf2013..593d5a957b69 100644
@@ -34,7 +34,7 @@ static int riscv_clock_next_event(unsigned long delta,
 static unsigned int riscv_clock_event_irq;
 static DEFINE_PER_CPU(struct clock_event_device, riscv_clock_event) = {
 	.name			= "riscv_timer_clockevent",
-	.features		= CLOCK_EVT_FEAT_ONESHOT,
+	.features		= CLOCK_EVT_FEAT_ONESHOT | CLOCK_EVT_FEAT_C3STOP,
 	.rating			= 100,
 	.set_next_event		= riscv_clock_next_event,
 };
Signature · 1 line
2.35.1
author maintainer
#52 lore raw

PATCH 52/298 ARM: dts: aspeed: ast2600-evb: Enable RX delay for MAC0/MAC1

In reply to Greg Kroah-Hartman

From: Howard Chiu <[email protected]>

[ Upstream commit 4d338ee40ba89e508c5d3e1b4af956af7cb5e12e ]

Since mac0/1 and mac2/3 are physically located on different die, they have different properties by nature, which is mac0/1 has smaller delay step.

The property 'phy-mode' on ast2600 mac0 and mac1 is recommended to set to 'rgmii-rxid' which enables the RX interface delay from the PHY chip.
Refer page 45 of SDK User Guide v08.00
https://github.com/AspeedTech-BMC/openbmc/releases/download/v08.00/SDK_User_Guide_v08.00.pdf

---
 arch/arm/boot/dts/aspeed-ast2600-evb.dts | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm/boot/dts/aspeed-ast2600-evb.dts +2 -2
index b7eb552640cb..788448cdd6b3 100644
@@ -103,7 +103,7 @@
 &mac0 {
 	status = "okay";
 
-	phy-mode = "rgmii";
+	phy-mode = "rgmii-rxid";
 	phy-handle = <&ethphy0>;
 
 	pinctrl-names = "default";
@@ -114,7 +114,7 @@
 &mac1 {
 	status = "okay";
 
-	phy-mode = "rgmii";
+	phy-mode = "rgmii-rxid";
 	phy-handle = <&ethphy1>;
 
 	pinctrl-names = "default";
Signature · 1 line
2.35.1
author maintainer
#53 lore raw

PATCH 53/298 rtc: mt6397: check return value after calling platform_get_resource()

In reply to Greg Kroah-Hartman

From: Yang Yingliang <[email protected]>

[ Upstream commit d3b43eb505bffb8e4cdf6800c15660c001553fe6 ]

It will cause null-ptr-deref if platform_get_resource() returns NULL, we need check the return value.

---
 drivers/rtc/rtc-mt6397.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/rtc/rtc-mt6397.c +2 -0
index 80dc479a6ff0..1d297af80f87 100644
@@ -269,6 +269,8 @@ static int mtk_rtc_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -EINVAL;
 	rtc->addr_base = res->start;
 
 	rtc->data = of_device_get_match_data(&pdev->dev);
Signature · 1 line
2.35.1
author maintainer
#54 lore raw

PATCH 54/298 rtc: ftrtc010: Fix error handling in ftrtc010_rtc_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit b520cbe5be37b1b9b401c0b6ecbdae32575273db ]

In the error handling path, the clk_prepare_enable() function call should be balanced by a corresponding 'clk_disable_unprepare()' call , as already done in the remove function.

clk_disable_unprepare calls clk_disable() and clk_unprepare(). They will use IS_ERR_OR_NULL to check the argument.

---
 drivers/rtc/rtc-ftrtc010.c | 34 ++++++++++++++++++++++++----------
 1 file changed, 24 insertions(+), 10 deletions(-)
drivers/rtc/rtc-ftrtc010.c +24 -10
index 53bb08fe1cd4..25c6e7d9570f 100644
@@ -137,26 +137,34 @@ static int ftrtc010_rtc_probe(struct platform_device *pdev)
 		ret = clk_prepare_enable(rtc->extclk);
 		if (ret) {
 			dev_err(dev, "failed to enable EXTCLK\n");
-			return ret;
+			goto err_disable_pclk;
 		}
 	}
 
 	rtc->rtc_irq = platform_get_irq(pdev, 0);
-	if (rtc->rtc_irq < 0)
-		return rtc->rtc_irq;
+	if (rtc->rtc_irq < 0) {
+		ret = rtc->rtc_irq;
+		goto err_disable_extclk;
+	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!res)
-		return -ENODEV;
+	if (!res) {
+		ret = -ENODEV;
+		goto err_disable_extclk;
+	}
 
 	rtc->rtc_base = devm_ioremap(dev, res->start,
 				     resource_size(res));
-	if (!rtc->rtc_base)
-		return -ENOMEM;
+	if (!rtc->rtc_base) {
+		ret = -ENOMEM;
+		goto err_disable_extclk;
+	}
 
 	rtc->rtc_dev = devm_rtc_allocate_device(dev);
-	if (IS_ERR(rtc->rtc_dev))
-		return PTR_ERR(rtc->rtc_dev);
+	if (IS_ERR(rtc->rtc_dev)) {
+		ret = PTR_ERR(rtc->rtc_dev);
+		goto err_disable_extclk;
+	}
 
 	rtc->rtc_dev->ops = &ftrtc010_rtc_ops;
 
@@ -172,9 +180,15 @@ static int ftrtc010_rtc_probe(struct platform_device *pdev)
 	ret = devm_request_irq(dev, rtc->rtc_irq, ftrtc010_rtc_interrupt,
 			       IRQF_SHARED, pdev->name, dev);
 	if (unlikely(ret))
-		return ret;
+		goto err_disable_extclk;
 
 	return devm_rtc_register_device(rtc->rtc_dev);
+
+err_disable_extclk:
+	clk_disable_unprepare(rtc->extclk);
+err_disable_pclk:
+	clk_disable_unprepare(rtc->pclk);
+	return ret;
 }
 
 static int ftrtc010_rtc_remove(struct platform_device *pdev)
Signature · 1 line
2.35.1
author maintainer
#55 lore raw

PATCH 55/298 staging: r8188eu: add check for kzalloc

In reply to Greg Kroah-Hartman

From: Jiasheng Jiang <[email protected]>

[ Upstream commit f94b47c6bde624d6c07f43054087607c52054a95 ]

As kzalloc() may return null pointer, it should be better to check the return value and return error if fails in order to avoid dereference of null pointer.
Moreover, the return value of rtw_alloc_hwxmits() should also be dealt with.

Fixes 15865124feed ("staging: r8188eu: introduce new core dir for RTL8188eu driver")
---
 drivers/staging/r8188eu/core/rtw_xmit.c    | 13 +++++++++++--
 drivers/staging/r8188eu/include/rtw_xmit.h |  2 +-
 2 files changed, 12 insertions(+), 3 deletions(-)
drivers/staging/r8188eu/core/rtw_xmit.c +11 -2
index 8503059edc46..f4e9f6102539 100644
@@ -179,7 +179,12 @@ s32	_rtw_init_xmit_priv(struct xmit_priv *pxmitpriv, struct adapter *padapter)
 
 	pxmitpriv->free_xmit_extbuf_cnt = num_xmit_extbuf;
 
-	rtw_alloc_hwxmits(padapter);
+	res = rtw_alloc_hwxmits(padapter);
+	if (res) {
+		res = _FAIL;
+		goto exit;
+	}
+
 	rtw_init_hwxmits(pxmitpriv->hwxmits, pxmitpriv->hwxmit_entry);
 
 	for (i = 0; i < 4; i++)
@@ -1496,7 +1501,7 @@ s32 rtw_xmit_classifier(struct adapter *padapter, struct xmit_frame *pxmitframe)
 	return res;
 }
 
-void rtw_alloc_hwxmits(struct adapter *padapter)
+int rtw_alloc_hwxmits(struct adapter *padapter)
 {
 	struct hw_xmit *hwxmits;
 	struct xmit_priv *pxmitpriv = &padapter->xmitpriv;
@@ -1504,6 +1509,8 @@ void rtw_alloc_hwxmits(struct adapter *padapter)
 	pxmitpriv->hwxmit_entry = HWXMIT_ENTRY;
 
 	pxmitpriv->hwxmits = kzalloc(sizeof(struct hw_xmit) * pxmitpriv->hwxmit_entry, GFP_KERNEL);
+	if (!pxmitpriv->hwxmits)
+		return -ENOMEM;
 
 	hwxmits = pxmitpriv->hwxmits;
 
@@ -1520,6 +1527,8 @@ void rtw_alloc_hwxmits(struct adapter *padapter)
 		hwxmits[3] .sta_queue = &pxmitpriv->bk_pending;
 	} else {
 	}
+
+	return 0;
 }
 
 void rtw_free_hwxmits(struct adapter *padapter)
drivers/staging/r8188eu/include/rtw_xmit.h +1 -1
index b2df1480d66b..e73632972900 100644
@@ -341,7 +341,7 @@ s32 rtw_txframes_sta_ac_pending(struct adapter *padapter,
 void rtw_init_hwxmits(struct hw_xmit *phwxmit, int entry);
 s32 _rtw_init_xmit_priv(struct xmit_priv *pxmitpriv, struct adapter *padapter);
 void _rtw_free_xmit_priv(struct xmit_priv *pxmitpriv);
-void rtw_alloc_hwxmits(struct adapter *padapter);
+int rtw_alloc_hwxmits(struct adapter *padapter);
 void rtw_free_hwxmits(struct adapter *padapter);
 s32 rtw_xmit(struct adapter *padapter, struct sk_buff **pkt);
 
Signature · 1 line
2.35.1
author maintainer
#56 lore raw

PATCH 56/298 serial: meson: acquire port->lock in startup()

In reply to Greg Kroah-Hartman

From: John Ogness <[email protected]>

[ Upstream commit 589f892ac8ef244e47c5a00ffd8605daa1eaef8e ]

The uart_ops startup() callback is called without interrupts disabled and without port->lock locked, relatively late during the boot process (from the call path of console_on_rootfs()). If the device is a console, it was already previously registered and could be actively printing messages.

Since the startup() callback is reading/writing registers used by the console write() callback (AML_UART_CONTROL), its access must be synchronized using the port->lock. Currently it is not.

The startup() callback is the only function that explicitly enables interrupts. Without the synchronization, it is possible that interrupts become accidentally permanently disabled.

CPU0                           CPU1
meson_serial_console_write     meson_uart_startup
--------------------------     ------------------

spin_lock(port->lock)
val = readl(AML_UART_CONTROL)
uart_console_write()
writel(INT_EN, AML_UART_CONTROL)
writel(val, AML_UART_CONTROL)
spin_unlock(port->lock)

Add port->lock synchronization to meson_uart_startup() to avoid racing with meson_serial_console_write().

Also add detailed comments to meson_uart_reset() explaining why it is *not* using port->lock synchronization.

---
 drivers/tty/serial/meson_uart.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
drivers/tty/serial/meson_uart.c +13 -0
index 45e00d928253..54a6b488bc8c 100644
@@ -253,6 +253,14 @@ static const char *meson_uart_type(struct uart_port *port)
 	return (port->type == PORT_MESON) ? "meson_uart" : NULL;
 }
 
+/*
+ * This function is called only from probe() using a temporary io mapping
+ * in order to perform a reset before setting up the device. Since the
+ * temporarily mapped region was successfully requested, there can be no
+ * console on this port at this time. Hence it is not necessary for this
+ * function to acquire the port->lock. (Since there is no console on this
+ * port at this time, the port->lock is not initialized yet.)
+ */
 static void meson_uart_reset(struct uart_port *port)
 {
 	u32 val;
@@ -267,9 +275,12 @@ static void meson_uart_reset(struct uart_port *port)
 
 static int meson_uart_startup(struct uart_port *port)
 {
+	unsigned long flags;
 	u32 val;
 	int ret = 0;
 
+	spin_lock_irqsave(&port->lock, flags);
+
 	val = readl(port->membase + AML_UART_CONTROL);
 	val |= AML_UART_CLEAR_ERR;
 	writel(val, port->membase + AML_UART_CONTROL);
@@ -285,6 +296,8 @@ static int meson_uart_startup(struct uart_port *port)
 	val = (AML_UART_RECV_IRQ(1) | AML_UART_XMIT_IRQ(port->fifosize / 2));
 	writel(val, port->membase + AML_UART_MISC);
 
+	spin_unlock_irqrestore(&port->lock, flags);
+
 	ret = request_irq(port->irq, meson_uart_interrupt, 0,
 			  port->name, port);
 
Signature · 1 line
2.35.1
author maintainer
#57 lore raw

PATCH 57/298 serial: 8250_fintek: Check SER_RS485_RTS_* only with RS485

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit af0179270977508df6986b51242825d7edd59caf ]

SER_RS485_RTS_ON_SEND and SER_RS485_RTS_AFTER_SEND relate to behavior within RS485 operation. The driver checks if they have the same value which is not possible to realize with the hardware. The check is taken regardless of SER_RS485_ENABLED flag and -EINVAL is returned when the check fails, which creates problems.

This check makes it unnecessarily complicated to turn RS485 mode off as simple zeroed serial_rs485 struct will trigger that equal values check. In addition, the driver itself memsets its rs485 structure to zero when RS485 is disabled but if userspace would try to make an TIOCSRS485 ioctl() call with the very same struct, it would end up failing with -EINVAL which doesn't make much sense.

Resolve the problem by moving the check inside SER_RS485_ENABLED block.

Fixes 7ecc77011c6f ("serial: 8250_fintek: Return -EINVAL on invalid configuration")
Cc Ricardo Ribalda Delgado <[email protected]>
---
 drivers/tty/serial/8250/8250_fintek.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/tty/serial/8250/8250_fintek.c +4 -4
index 251f0018ae8c..dba5950b8d0e 100644
@@ -200,12 +200,12 @@ static int fintek_8250_rs485_config(struct uart_port *port,
 	if (!pdata)
 		return -EINVAL;
 
-	/* Hardware do not support same RTS level on send and receive */
-	if (!(rs485->flags & SER_RS485_RTS_ON_SEND) ==
-			!(rs485->flags & SER_RS485_RTS_AFTER_SEND))
-		return -EINVAL;
 
 	if (rs485->flags & SER_RS485_ENABLED) {
+		/* Hardware do not support same RTS level on send and receive */
+		if (!(rs485->flags & SER_RS485_RTS_ON_SEND) ==
+		    !(rs485->flags & SER_RS485_RTS_AFTER_SEND))
+			return -EINVAL;
 		memset(rs485->padding, 0, sizeof(rs485->padding));
 		config |= RS485_URA;
 	} else {
Signature · 1 line
2.35.1
author maintainer
#58 lore raw

PATCH 58/298 serial: cpm_uart: Fix build error without CONFIG_SERIAL_CPM_CONSOLE

In reply to Greg Kroah-Hartman

From: YueHaibing <[email protected]>

[ Upstream commit 0258502f11a4f6036b5f8b34b09027c8a92def3a ]

drivers/tty/serial/cpm_uart/cpm_uart_core.c: In function ‘cpm_uart_init_port’:
drivers/tty/serial/cpm_uart/cpm_uart_core.c:1251:7: error: ‘udbg_port’ undeclared (first use in this function); did you mean ‘uart_port’?
if (!udbg_port)

       ^~~~~~~~~
       uart_port

commit d142585bceb3 leave this corner, wrap it with #ifdef block

Fixes d142585bceb3 ("serial: cpm_uart: Protect udbg definitions by CONFIG_SERIAL_CPM_CONSOLE")
---
 drivers/tty/serial/cpm_uart/cpm_uart_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/tty/serial/cpm_uart/cpm_uart_core.c +1 -1
index d6d3db9c3b1f..db07d6a5d764 100644
@@ -1247,7 +1247,7 @@ static int cpm_uart_init_port(struct device_node *np,
 	}
 
 #ifdef CONFIG_PPC_EARLY_DEBUG_CPM
-#ifdef CONFIG_CONSOLE_POLL
+#if defined(CONFIG_CONSOLE_POLL) && defined(CONFIG_SERIAL_CPM_CONSOLE)
 	if (!udbg_port)
 #endif
 		udbg_putc = NULL;
Signature · 1 line
2.35.1
author maintainer
#59 lore raw

PATCH 59/298 serial: uartlite: Fix BRKINT clearing

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 3f7fed405c118607d4d42255f2572072db728399 ]

BRKINT is within c_iflag rather than c_cflag.

Fixes ea017f5853e9 (tty: serial: uartlite: Prevent changing fixed parameters)
---
 drivers/tty/serial/uartlite.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/tty/serial/uartlite.c +2 -1
index e1fa52d31474..7c788c697f3e 100644
@@ -321,7 +321,8 @@ static void ulite_set_termios(struct uart_port *port, struct ktermios *termios,
 	struct uartlite_data *pdata = port->private_data;
 
 	/* Set termios to what the hardware supports */
-	termios->c_cflag &= ~(BRKINT | CSTOPB | PARENB | PARODD | CSIZE);
+	termios->c_iflag &= ~BRKINT;
+	termios->c_cflag &= ~(CSTOPB | PARENB | PARODD | CSIZE);
 	termios->c_cflag |= pdata->cflags & (PARENB | PARODD | CSIZE);
 	tty_termios_encode_baud_rate(termios, pdata->baud, pdata->baud);
 
Signature · 1 line
2.35.1
author maintainer
#60 lore raw

PATCH 60/298 serial: digicolor-usart: Dont allow CS5-6

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit fd63031b8c0763addcecdefe0e0c59d49646204e ]

Only CS7 and CS8 seem supported but CSIZE is not sanitized to CS8 in the default: block.

Set CSIZE correctly so that userspace knows the effective value. Incorrect CSIZE also results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

---
 drivers/tty/serial/digicolor-usart.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/tty/serial/digicolor-usart.c +2 -0
index c7f81aa1ce91..5fea9bf86e85 100644
@@ -309,6 +309,8 @@ static void digicolor_uart_set_termios(struct uart_port *port,
 	case CS8:
 	default:
 		config |= UA_CONFIG_CHAR_LEN;
+		termios->c_cflag &= ~CSIZE;
+		termios->c_cflag |= CS8;
 		break;
 	}
 
Signature · 1 line
2.35.1
author maintainer
#61 lore raw

PATCH 61/298 serial: rda-uart: Dont allow CS5-6

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 098333a9c7d12bb3ce44c82f08b4d810c44d31b0 ]

Only CS7 and CS8 are supported but CSIZE is not sanitized after fallthrough from CS5 or CS6 to CS7.

Set CSIZE correctly so that userspace knows the effective value. Incorrect CSIZE also results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

Fixes c10b13325ced (tty: serial: Add RDA8810PL UART driver)
Cc Manivannan Sadhasivam <[email protected]>
---
 drivers/tty/serial/rda-uart.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/tty/serial/rda-uart.c +2 -0
index d550d8fa2fab..a8fe1c3ebcd9 100644
@@ -262,6 +262,8 @@ static void rda_uart_set_termios(struct uart_port *port,
 		fallthrough;
 	case CS7:
 		ctrl &= ~RDA_UART_DBITS_8;
+		termios->c_cflag &= ~CSIZE;
+		termios->c_cflag |= CS7;
 		break;
 	default:
 		ctrl |= RDA_UART_DBITS_8;
Signature · 1 line
2.35.1
author maintainer
#62 lore raw

PATCH 62/298 serial: txx9: Dont allow CS5-6

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 79ac88655dc0551e3571ad16bdabdbe65d61553e ]

Only CS7 and CS8 are supported but CSIZE is not sanitized with CS5 or CS6 to CS8.

Set CSIZE correctly so that userspace knows the effective value. Incorrect CSIZE also results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

---
 drivers/tty/serial/serial_txx9.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/tty/serial/serial_txx9.c +2 -0
index aaca4fe38486..1f8362d5e3b9 100644
@@ -644,6 +644,8 @@ serial_txx9_set_termios(struct uart_port *port, struct ktermios *termios,
 	case CS6:	/* not supported */
 	case CS8:
 		cval |= TXX9_SILCR_UMODE_8BIT;
+		termios->c_cflag &= ~CSIZE;
+		termios->c_cflag |= CS8;
 		break;
 	}
 
Signature · 1 line
2.35.1
author maintainer
#63 lore raw

PATCH 63/298 serial: sh-sci: Dont allow CS5-6

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 9b87162de8be26bf3156460b37deee6399fd0fcb ]

Only CS7 and CS8 seem supported but CSIZE is not sanitized from CS5 or CS6 to CS8.

Set CSIZE correctly so that userspace knows the effective value. Incorrect CSIZE also results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

---
 drivers/tty/serial/sh-sci.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/tty/serial/sh-sci.c +5 -1
index 968967d722d4..e55895f0a4ff 100644
@@ -2379,8 +2379,12 @@ static void sci_set_termios(struct uart_port *port, struct ktermios *termios,
 	int best_clk = -1;
 	unsigned long flags;
 
-	if ((termios->c_cflag & CSIZE) == CS7)
+	if ((termios->c_cflag & CSIZE) == CS7) {
 		smr_val |= SCSMR_CHR;
+	} else {
+		termios->c_cflag &= ~CSIZE;
+		termios->c_cflag |= CS8;
+	}
 	if (termios->c_cflag & PARENB)
 		smr_val |= SCSMR_PE;
 	if (termios->c_cflag & PARODD)
Signature · 1 line
2.35.1
author maintainer
#64 lore raw

PATCH 64/298 serial: sifive: Sanitize CSIZE and c_iflag

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit c069d2756c01ed36121fae6a42c14fdf1325c71d ]

Only CS8 is supported but CSIZE was not sanitized to CS8.

Set CSIZE correctly so that userspace knows the effective value. Incorrect CSIZE also results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

Similarly, INPCK, PARMRK, and BRKINT are reported textually unsupported but were not cleared in termios c_iflag which is the machine-readable format.

Fixes 45c054d0815b (tty: serial: add driver for the SiFive UART)
Cc Paul Walmsley <[email protected]>
---
 drivers/tty/serial/sifive.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/tty/serial/sifive.c +5 -1
index fff0b7916f3d..cba44483cb03 100644
@@ -666,12 +666,16 @@ static void sifive_serial_set_termios(struct uart_port *port,
 	int rate;
 	char nstop;
 
-	if ((termios->c_cflag & CSIZE) != CS8)
+	if ((termios->c_cflag & CSIZE) != CS8) {
 		dev_err_once(ssp->port.dev, "only 8-bit words supported\n");
+		termios->c_cflag &= ~CSIZE;
+		termios->c_cflag |= CS8;
+	}
 	if (termios->c_iflag & (INPCK | PARMRK))
 		dev_err_once(ssp->port.dev, "parity checking not supported\n");
 	if (termios->c_iflag & BRKINT)
 		dev_err_once(ssp->port.dev, "BREAK detection not supported\n");
+	termios->c_iflag &= ~(INPCK|PARMRK|BRKINT);
 
 	/* Set number of stop bits */
 	nstop = (termios->c_cflag & CSTOPB) ? 2 : 1;
Signature · 1 line
2.35.1
author maintainer
#65 lore raw

PATCH 65/298 serial: st-asc: Sanitize CSIZE and correct PARENB for CS7

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 52bb1cb7118564166b04d52387bd8403632f5190 ]

Only CS7 and CS8 seem supported but CSIZE is not sanitized from CS5 or CS6 to CS8. In addition, ASC_CTL_MODE_7BIT_PAR suggests that CS7 has to have parity, thus add PARENB.

Incorrect CSIZE results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

Fixes c4b058560762 (serial:st-asc: Add ST ASC driver.)
Cc Srinivas Kandagatla <[email protected]>
---
 drivers/tty/serial/st-asc.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/tty/serial/st-asc.c +4 -0
index 87e480cc8206..5a45633aaea8 100644
@@ -535,10 +535,14 @@ static void asc_set_termios(struct uart_port *port, struct ktermios *termios,
 	/* set character length */
 	if ((cflag & CSIZE) == CS7) {
 		ctrl_val |= ASC_CTL_MODE_7BIT_PAR;
+		cflag |= PARENB;
 	} else {
 		ctrl_val |= (cflag & PARENB) ?  ASC_CTL_MODE_8BIT_PAR :
 						ASC_CTL_MODE_8BIT;
+		cflag &= ~CSIZE;
+		cflag |= CS8;
 	}
+	termios->c_cflag = cflag;
 
 	/* set stop bit */
 	ctrl_val |= (cflag & CSTOPB) ? ASC_CTL_STOP_2BIT : ASC_CTL_STOP_1BIT;
Signature · 1 line
2.35.1
author maintainer
#66 lore raw

PATCH 66/298 serial: stm32-usart: Correct CSIZE, bits, and parity

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 1deeda8d2877c18bc2b9eeee10dd6d2628852848 ]

Add CSIZE sanitization for unsupported CSIZE configurations. In addition, if parity is asked for but CSx was unsupported, the sensible result is CS8+parity which requires setting USART_CR1_M0 like with 9 bits.

Incorrect CSIZE results in miscalculation of the frame bits in tty_get_char_size() or in its predecessor where the roughly the same code is directly within uart_update_timeout().

Fixes c8a9d043947b (serial: stm32: fix word length configuration)
Cc Erwan Le Ray <[email protected]>
---
 drivers/tty/serial/stm32-usart.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)
drivers/tty/serial/stm32-usart.c +12 -3
index 9570002d07e7..9bc970be59ba 100644
@@ -1037,13 +1037,22 @@ static void stm32_usart_set_termios(struct uart_port *port,
 	 * CS8 or (CS7 + parity), 8 bits word aka [M1:M0] = 0b00
 	 * M0 and M1 already cleared by cr1 initialization.
 	 */
-	if (bits == 9)
+	if (bits == 9) {
 		cr1 |= USART_CR1_M0;
-	else if ((bits == 7) && cfg->has_7bits_data)
+	} else if ((bits == 7) && cfg->has_7bits_data) {
 		cr1 |= USART_CR1_M1;
-	else if (bits != 8)
+	} else if (bits != 8) {
 		dev_dbg(port->dev, "Unsupported data bits config: %u bits\n"
 			, bits);
+		cflag &= ~CSIZE;
+		cflag |= CS8;
+		termios->c_cflag = cflag;
+		bits = 8;
+		if (cflag & PARENB) {
+			bits++;
+			cr1 |= USART_CR1_M0;
+		}
+	}
 
 	if (ofs->rtor != UNDEF_REG && (stm32_port->rx_ch ||
 				       (stm32_port->fifoen &&
Signature · 1 line
2.35.1
author maintainer
#67 lore raw

PATCH 67/298 firmware: dmi-sysfs: Fix memory leak in dmi_sysfs_register_handle

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 660ba678f9998aca6db74f2dd912fa5124f0fa31 ]

kobject_init_and_add() takes reference even when it fails. According to the doc of kobject_init_and_add()

If this function returns an error, kobject_put() must be called to properly clean up the memory associated with the object.

Fix this issue by calling kobject_put().

---
 drivers/firmware/dmi-sysfs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/firmware/dmi-sysfs.c +1 -1
index 3a353776bd34..66727ad3361b 100644
@@ -604,7 +604,7 @@ static void __init dmi_sysfs_register_handle(const struct dmi_header *dh,
 				    "%d-%d", dh->type, entry->instance);
 
 	if (*ret) {
-		kfree(entry);
+		kobject_put(&entry->kobj);
 		return;
 	}
 
Signature · 1 line
2.35.1
author maintainer
#68 lore raw

PATCH 68/298 bus: ti-sysc: Fix warnings for unbind for serial

In reply to Greg Kroah-Hartman

From: Tony Lindgren <[email protected]>

[ Upstream commit c337125b8834f9719dfda0e40b25eaa266f1b8cf ]

We can get "failed to disable" clock_unprepare warnings on unbind at least for the serial console device if the unbind is done before the device has been idled.

As some devices are using deferred idle, we must check the status for pending idle work to idle the device.

Fixes 76f0f772e469 ("bus: ti-sysc: Improve handling for no-reset-on-init and no-idle-on-init")
Cc Romain Naour <[email protected]>
---
 drivers/bus/ti-sysc.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/bus/ti-sysc.c +3 -1
index 7a1b1f9e4933..70d00cea9d22 100644
@@ -3395,7 +3395,9 @@ static int sysc_remove(struct platform_device *pdev)
 	struct sysc *ddata = platform_get_drvdata(pdev);
 	int error;
 
-	cancel_delayed_work_sync(&ddata->idle_work);
+	/* Device can still be enabled, see deferred idle quirk in probe */
+	if (cancel_delayed_work_sync(&ddata->idle_work))
+		ti_sysc_idle(&ddata->idle_work.work);
 
 	error = pm_runtime_resume_and_get(ddata->dev);
 	if (error < 0) {
Signature · 1 line
2.35.1
author maintainer
#69 lore raw

PATCH 69/298 driver: base: fix UAF when driver_attach failed

In reply to Greg Kroah-Hartman

From: Schspa Shi <[email protected]>

[ Upstream commit 310862e574001a97ad02272bac0fd13f75f42a27 ]

When driver_attach(drv); failed, the driver_private will be freed. But it has been added to the bus, which caused a UAF.

To fix it, we need to delete it from the bus when failed.

Fixes 190888ac01d0 ("driver core: fix possible missing of device probe")
---
 drivers/base/bus.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/base/bus.c +3 -1
index 97936ec49bde..7ca47e5b3c1f 100644
@@ -617,7 +617,7 @@ int bus_add_driver(struct device_driver *drv)
 	if (drv->bus->p->drivers_autoprobe) {
 		error = driver_attach(drv);
 		if (error)
-			goto out_unregister;
+			goto out_del_list;
 	}
 	module_add_driver(drv->owner, drv);
 
@@ -644,6 +644,8 @@ int bus_add_driver(struct device_driver *drv)
 
 	return 0;
 
+out_del_list:
+	klist_del(&priv->knode_bus);
 out_unregister:
 	kobject_put(&priv->kobj);
 	/* drv->p is freed in driver_release()  */
Signature · 1 line
2.35.1
author maintainer
#70 lore raw

PATCH 70/298 driver core: fix deadlock in __device_attach

In reply to Greg Kroah-Hartman

From: Zhang Wensheng <[email protected]>

[ Upstream commit b232b02bf3c205b13a26dcec08e53baddd8e59ed ]

In __device_attach function, The lock holding logic is as follows:
...
__device_attach
device_lock(dev) // get lock dev
async_schedule_dev(__device_attach_async_helper, dev); // func

    async_schedule_node
      async_schedule_node_domain(func)
        entry = kzalloc(sizeof(struct async_entry), GFP_ATOMIC);
	/* when fail or work limit, sync to execute func, but
	   __device_attach_async_helper will get lock dev as
	   well, which will lead to A-A deadlock.  */
	if (!entry || atomic_read(&entry_count) > MAX_WORK) {
	  func;
	else
	  queue_work_node(node, system_unbound_wq, &entry->work)

device_unlock(dev)

As shown above, when it is allowed to do async probes, because of out of memory or work limit, async work is not allowed, to do sync execute instead. it will lead to A-A deadlock because of __device_attach_async_helper getting lock dev.

To fix the deadlock, move the async_schedule_dev outside device_lock, as we can see, in async_schedule_node_domain, the parameter of queue_work_node is system_unbound_wq, so it can accept concurrent operations. which will also not change the code logic, and will not lead to deadlock.

Fixes 765230b5f084 ("driver-core: add asynchronous probing support for drivers")
---
 drivers/base/dd.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/base/dd.c +4 -1
index 7e079fa3795b..977e94cf669e 100644
@@ -945,6 +945,7 @@ static void __device_attach_async_helper(void *_dev, async_cookie_t cookie)
 static int __device_attach(struct device *dev, bool allow_async)
 {
 	int ret = 0;
+	bool async = false;
 
 	device_lock(dev);
 	if (dev->p->dead) {
@@ -983,7 +984,7 @@ static int __device_attach(struct device *dev, bool allow_async)
 			 */
 			dev_dbg(dev, "scheduling asynchronous probe\n");
 			get_device(dev);
-			async_schedule_dev(__device_attach_async_helper, dev);
+			async = true;
 		} else {
 			pm_request_idle(dev);
 		}
@@ -993,6 +994,8 @@ static int __device_attach(struct device *dev, bool allow_async)
 	}
 out_unlock:
 	device_unlock(dev);
+	if (async)
+		async_schedule_dev(__device_attach_async_helper, dev);
 	return ret;
 }
 
Signature · 1 line
2.35.1
author maintainer
#71 lore raw

PATCH 71/298 watchdog: rti-wdt: Fix pm_runtime_get_sync() error checking

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit b3ac0c58fa8934926360268f3d89ec7680644d7b ]

If the device is already in a runtime PM enabled state pm_runtime_get_sync() will return 1, so a test for negative value should be used to check for errors.

---
 drivers/watchdog/rti_wdt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/watchdog/rti_wdt.c +1 -1
index db843f825860..00ebeffc674f 100644
@@ -226,7 +226,7 @@ static int rti_wdt_probe(struct platform_device *pdev)
 
 	pm_runtime_enable(dev);
 	ret = pm_runtime_get_sync(dev);
-	if (ret) {
+	if (ret < 0) {
 		pm_runtime_put_noidle(dev);
 		pm_runtime_disable(&pdev->dev);
 		return dev_err_probe(dev, ret, "runtime pm failed\n");
Signature · 1 line
2.35.1
author maintainer
#72 lore raw

PATCH 72/298 watchdog: ts4800_wdt: Fix refcount leak in ts4800_wdt_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 5d24df3d690809952528e7a19a43d84bc5b99d44 ]

of_parse_phandle() returns a node pointer with refcount incremented, we should use of_node_put() on it when done. Add missing of_node_put() in some error paths.

---
 drivers/watchdog/ts4800_wdt.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/watchdog/ts4800_wdt.c +4 -1
index c137ad2bd5c3..0ea554c7cda5 100644
@@ -125,13 +125,16 @@ static int ts4800_wdt_probe(struct platform_device *pdev)
 	ret = of_property_read_u32_index(np, "syscon", 1, &reg);
 	if (ret < 0) {
 		dev_err(dev, "no offset in syscon\n");
+		of_node_put(syscon_np);
 		return ret;
 	}
 
 	/* allocate memory for watchdog struct */
 	wdt = devm_kzalloc(dev, sizeof(*wdt), GFP_KERNEL);
-	if (!wdt)
+	if (!wdt) {
+		of_node_put(syscon_np);
 		return -ENOMEM;
+	}
 
 	/* set regmap and offset to know where to write */
 	wdt->feed_offset = reg;
Signature · 1 line
2.35.1
author maintainer
#73 lore raw

PATCH 73/298 blk-mq: dont touch ->tagset in blk_mq_get_sq_hctx

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit 5d05426e2d5fd7df8afc866b78c36b37b00188b7 ]

blk_mq_run_hw_queues() could be run when there isn't queued request and after queue is cleaned up, at that time tagset is freed, because tagset lifetime is covered by driver, and often freed after blk_cleanup_queue() returns.

So don't touch ->tagset for figuring out current default hctx by the mapping built in request queue, so use-after-free on tagset can be avoided. Meantime this way should be fast than retrieving mapping from tagset.

Cc "yukuai (C)" <[email protected]>
Cc Jan Kara <[email protected]>
Fixes b6e68ee82585 ("blk-mq: Improve performance of non-mq IO schedulers with multiple HW queues")
---
 block/blk-mq.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
block/blk-mq.c +3 -4
index 0aa20df31e36..6afe0cd128ac 100644
@@ -2114,8 +2114,7 @@ static bool blk_mq_has_sqsched(struct request_queue *q)
  */
 static struct blk_mq_hw_ctx *blk_mq_get_sq_hctx(struct request_queue *q)
 {
-	struct blk_mq_hw_ctx *hctx;
-
+	struct blk_mq_ctx *ctx = blk_mq_get_ctx(q);
 	/*
 	 * If the IO scheduler does not respect hardware queues when
 	 * dispatching, we just don't bother with multiple HW queues and
@@ -2123,8 +2122,8 @@ static struct blk_mq_hw_ctx *blk_mq_get_sq_hctx(struct request_queue *q)
 	 * just causes lock contention inside the scheduler and pointless cache
 	 * bouncing.
 	 */
-	hctx = blk_mq_map_queue_type(q, HCTX_TYPE_DEFAULT,
-				     raw_smp_processor_id());
+	struct blk_mq_hw_ctx *hctx = blk_mq_map_queue(q, 0, ctx);
+
 	if (!blk_mq_hctx_stopped(hctx))
 		return hctx;
 	return NULL;
Signature · 1 line
2.35.1
author maintainer
#74 lore raw

PATCH 74/298 ASoC: fsl_sai: Fix FSL_SAI_xDR/xFR definition

In reply to Greg Kroah-Hartman

From: Shengjiu Wang <[email protected]>

[ Upstream commit e4dd748dc87cf431af7b3954963be0d9f6150217 ]

There are multiple xDR and xFR registers, the index is from 0 to 7. FSL_SAI_xDR and FSL_SAI_xFR is abandoned, replace them with FSL_SAI_xDR0 and FSL_SAI_xFR0.

Fixes 4f7a0728b530 ("ASoC: fsl_sai: Add support for SAI new version")
---
 sound/soc/fsl/fsl_sai.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
sound/soc/fsl/fsl_sai.h +2 -2
index 9aaf231bc024..93da86009c75 100644
@@ -80,8 +80,8 @@
 #define FSL_SAI_xCR3(tx, ofs)	(tx ? FSL_SAI_TCR3(ofs) : FSL_SAI_RCR3(ofs))
 #define FSL_SAI_xCR4(tx, ofs)	(tx ? FSL_SAI_TCR4(ofs) : FSL_SAI_RCR4(ofs))
 #define FSL_SAI_xCR5(tx, ofs)	(tx ? FSL_SAI_TCR5(ofs) : FSL_SAI_RCR5(ofs))
-#define FSL_SAI_xDR(tx, ofs)	(tx ? FSL_SAI_TDR(ofs) : FSL_SAI_RDR(ofs))
-#define FSL_SAI_xFR(tx, ofs)	(tx ? FSL_SAI_TFR(ofs) : FSL_SAI_RFR(ofs))
+#define FSL_SAI_xDR0(tx)	(tx ? FSL_SAI_TDR0 : FSL_SAI_RDR0)
+#define FSL_SAI_xFR0(tx)	(tx ? FSL_SAI_TFR0 : FSL_SAI_RFR0)
 #define FSL_SAI_xMR(tx)		(tx ? FSL_SAI_TMR : FSL_SAI_RMR)
 
 /* SAI Transmit/Receive Control Register */
Signature · 1 line
2.35.1
author maintainer
#75 lore raw

PATCH 75/298 clocksource/drivers/oxnas-rps: Fix irq_of_parse_and_map() return value

In reply to Greg Kroah-Hartman

From: Krzysztof Kozlowski <[email protected]>

[ Upstream commit 9c04a8ff03def4df3f81219ffbe1ec9b44ff5348 ]

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

Fixes 89355274e1f7 ("clocksource/drivers/oxnas-rps: Add Oxford Semiconductor RPS Dual Timer")
---
 drivers/clocksource/timer-oxnas-rps.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clocksource/timer-oxnas-rps.c +1 -1
index 56c0cc32d0ac..d514b44e67dd 100644
@@ -236,7 +236,7 @@ static int __init oxnas_rps_timer_init(struct device_node *np)
 	}
 
 	rps->irq = irq_of_parse_and_map(np, 0);
-	if (rps->irq < 0) {
+	if (!rps->irq) {
 		ret = -EINVAL;
 		goto err_iomap;
 	}
Signature · 1 line
2.35.1
author maintainer
#76 lore raw

PATCH 76/298 s390/crypto: fix scatterwalk_unmap() callers in AES-GCM

In reply to Greg Kroah-Hartman

From: Jann Horn <[email protected]>

[ Upstream commit bd52cd5e23f134019b23f0c389db0f9a436e4576 ]

The argument of scatterwalk_unmap() is supposed to be the void* that was returned by the previous scatterwalk_map() call.
The s390 AES-GCM implementation was instead passing the pointer to the struct scatter_walk.

This doesn't actually break anything because scatterwalk_unmap() only uses its argument under CONFIG_HIGHMEM and ARCH_HAS_FLUSH_ON_KUNMAP.

Fixes bf7fa038707c ("s390/crypto: add s390 platform specific aes gcm support.")
---
 arch/s390/crypto/aes_s390.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/s390/crypto/aes_s390.c +2 -2
index 54c7536f2482..1023e9d43d44 100644
@@ -701,7 +701,7 @@ static inline void _gcm_sg_unmap_and_advance(struct gcm_sg_walk *gw,
 					     unsigned int nbytes)
 {
 	gw->walk_bytes_remain -= nbytes;
-	scatterwalk_unmap(&gw->walk);
+	scatterwalk_unmap(gw->walk_ptr);
 	scatterwalk_advance(&gw->walk, nbytes);
 	scatterwalk_done(&gw->walk, 0, gw->walk_bytes_remain);
 	gw->walk_ptr = NULL;
@@ -776,7 +776,7 @@ static int gcm_out_walk_go(struct gcm_sg_walk *gw, unsigned int minbytesneeded)
 		goto out;
 	}
 
-	scatterwalk_unmap(&gw->walk);
+	scatterwalk_unmap(gw->walk_ptr);
 	gw->walk_ptr = NULL;
 
 	gw->ptr = gw->buf;
Signature · 1 line
2.35.1
author maintainer
#77 lore raw

PATCH 77/298 amt: fix return value of amt_update_handler()

In reply to Greg Kroah-Hartman

From: Taehee Yoo <[email protected]>

[ Upstream commit ac1dbf55981b88d64312858ea06e3e63001f085d ]

If a relay receives an update message, it lookup a tunnel. and if there is no tunnel for that message, it should be treated as an error, not a success.
But amt_update_handler() returns false, which means success.

Fixes cbc21dc1cfe9 ("amt: add data plane of amt interface")
---
 drivers/net/amt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/amt.c +1 -1
index fb774d568baa..6205282a09e5 100644
@@ -2423,7 +2423,7 @@ static bool amt_update_handler(struct amt_dev *amt, struct sk_buff *skb)
 		}
 	}
 
-	return false;
+	return true;
 
 report:
 	iph = ip_hdr(skb);
Signature · 1 line
2.35.1
author maintainer
#78 lore raw

PATCH 78/298 amt: fix possible memory leak in amt_rcv()

In reply to Greg Kroah-Hartman

From: Taehee Yoo <[email protected]>

[ Upstream commit 1a1a0e80e005cbdc2c250fc858e1d8570f4e4acb ]

If an amt receives packets and it finds socket. If it can't find a socket, it should free a received skb. But it doesn't.
So, a memory leak would possibly occur.

Fixes cbc21dc1cfe9 ("amt: add data plane of amt interface")
---
 drivers/net/amt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/amt.c +1 -1
index 6205282a09e5..d23eac9ce858 100644
@@ -2679,7 +2679,7 @@ static int amt_rcv(struct sock *sk, struct sk_buff *skb)
 	amt = rcu_dereference_sk_user_data(sk);
 	if (!amt) {
 		err = true;
-		goto out;
+		goto drop;
 	}
 
 	skb->dev = amt->dev;
Signature · 1 line
2.35.1
author maintainer
#79 lore raw

PATCH 79/298 net/smc: set ini->smcrv2.ib_dev_v2 to NULL if SMC-Rv2 is unavailable

In reply to Greg Kroah-Hartman

From: liuyacan <[email protected]>

[ Upstream commit b3b1a17538d3ef6a9667b2271216fd16d7678ab5 ]

In the process of checking whether RDMAv2 is available, the current implementation first sets ini->smcrv2.ib_dev_v2, and then allocates smc buf desc and register rmb, but the latter may fail. In this case, the pointer should be reset.

Fixes e49300a6bf62 ("net/smc: add listen processing for SMC-Rv2")
---
 net/smc/af_smc.c | 1 +
 1 file changed, 1 insertion(+)
net/smc/af_smc.c +1 -0
index b9fe31834354..4bc6b16669f3 100644
@@ -1973,6 +1973,7 @@ static void smc_find_rdma_v2_device_serv(struct smc_sock *new_smc,
 
 not_found:
 	ini->smcr_version &= ~SMC_V2;
+	ini->smcrv2.ib_dev_v2 = NULL;
 	ini->check_smcrv2 = false;
 }
 
Signature · 1 line
2.35.1
author maintainer
#80 lore raw

PATCH 80/298 spi: fsi: Fix spurious timeout

In reply to Greg Kroah-Hartman

From: Eddie James <[email protected]>

[ Upstream commit 61bf40ef51aa73f6216b33563271b6acf7ea8d70 ]

The driver may return a timeout error even if the status register indicates that the transfer may proceed. Fix this by restructuring the polling loop.

Fixes 89b35e3f2851 ("spi: fsi: Implement a timeout for polling status")
---
 drivers/spi/spi-fsi.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/spi/spi-fsi.c +6 -6
index d403a7a3021d..72ab066ce552 100644
@@ -319,12 +319,12 @@ static int fsi_spi_transfer_data(struct fsi_spi *ctx,
 
 			end = jiffies + msecs_to_jiffies(SPI_FSI_STATUS_TIMEOUT_MS);
 			do {
+				if (time_after(jiffies, end))
+					return -ETIMEDOUT;
+
 				rc = fsi_spi_status(ctx, &status, "TX");
 				if (rc)
 					return rc;
-
-				if (time_after(jiffies, end))
-					return -ETIMEDOUT;
 			} while (status & SPI_FSI_STATUS_TDR_FULL);
 
 			sent += nb;
@@ -337,12 +337,12 @@ static int fsi_spi_transfer_data(struct fsi_spi *ctx,
 		while (transfer->len > recv) {
 			end = jiffies + msecs_to_jiffies(SPI_FSI_STATUS_TIMEOUT_MS);
 			do {
+				if (time_after(jiffies, end))
+					return -ETIMEDOUT;
+
 				rc = fsi_spi_status(ctx, &status, "RX");
 				if (rc)
 					return rc;
-
-				if (time_after(jiffies, end))
-					return -ETIMEDOUT;
 			} while (!(status & SPI_FSI_STATUS_RDR_FULL));
 
 			rc = fsi_spi_read_reg(ctx, SPI_FSI_DATA_RX, &in);
Signature · 1 line
2.35.1
author maintainer
#81 lore raw

PATCH 81/298 drm/amdgpu: Off by one in dm_dmub_outbox1_low_irq()

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit a35faec3db0e13aac8ea720bc1a3503081dd5a3d ]

The > ARRAY_SIZE() should be >= ARRAY_SIZE() to prevent an out of bounds access.

Fixes e27c41d5b068 ("drm/amd/display: Support for DMUB HPD interrupt handling")
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +1 -1
index 24db2297857b..edb5e72aeb66 100644
@@ -767,7 +767,7 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
 
 		do {
 			dc_stat_get_dmub_notification(adev->dm.dc, &notify);
-			if (notify.type > ARRAY_SIZE(dm->dmub_thread_offload)) {
+			if (notify.type >= ARRAY_SIZE(dm->dmub_thread_offload)) {
 				DRM_ERROR("DM: notify type %d invalid!", notify.type);
 				continue;
 			}
Signature · 1 line
2.35.1
author maintainer
#82 lore raw

PATCH 82/298 net: lan966x: check devm_of_phy_get() for -EDEFER_PROBE

In reply to Greg Kroah-Hartman

From: Michael Walle <[email protected]>

[ Upstream commit b58cdd4388b1d8f5bee9f5a3897a7e780d1eaa48 ]

At the moment, if devm_of_phy_get() returns an error the serdes simply isn't set. While it is bad to ignore an error in general, there is a particular bug that network isn't working if the serdes driver is compiled as a module. In that case, devm_of_phy_get() returns -EDEFER_PROBE and the error is silently ignored.

The serdes is optional, it is not there if the port is using RGMII, in which case devm_of_phy_get() returns -ENODEV. Rearrange the error handling so that -ENODEV will be handled but other error codes will abort the probing.

Fixes d28d6d2e37d1 ("net: lan966x: add port module support")
---
 drivers/net/ethernet/microchip/lan966x/lan966x_main.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/net/ethernet/microchip/lan966x/lan966x_main.c +7 -2
index fee148bbf13e..b0cb3b65cd5b 100644
@@ -948,8 +948,13 @@ static int lan966x_probe(struct platform_device *pdev)
 		lan966x->ports[p]->fwnode = fwnode_handle_get(portnp);
 
 		serdes = devm_of_phy_get(lan966x->dev, to_of_node(portnp), NULL);
-		if (!IS_ERR(serdes))
-			lan966x->ports[p]->serdes = serdes;
+		if (PTR_ERR(serdes) == -ENODEV)
+			serdes = NULL;
+		if (IS_ERR(serdes)) {
+			err = PTR_ERR(serdes);
+			goto cleanup_ports;
+		}
+		lan966x->ports[p]->serdes = serdes;
 
 		lan966x_port_init(lan966x->ports[p]);
 	}
Signature · 1 line
2.35.1
author maintainer
#83 lore raw

PATCH 83/298 net: sched: fixed barrier to prevent skbuff sticking in qdisc backlog

In reply to Greg Kroah-Hartman

From: Vincent Ray <[email protected]>

[ Upstream commit a54ce3703613e41fe1d98060b62ec09a3984dc28 ]

In qdisc_run_begin(), smp_mb__before_atomic() used before test_bit() does not provide any ordering guarantee as test_bit() is not an atomic operation. This, added to the fact that the spin_trylock() call at the beginning of qdisc_run_begin() does not guarantee acquire semantics if it does not grab the lock, makes it possible for the following statement :

if (test_bit(__QDISC_STATE_MISSED, &qdisc->state))

to be executed before an enqueue operation called before qdisc_run_begin().

As a result the following race can happen :

CPU 1 CPU 2

qdisc_run_begin() qdisc_run_begin() /* true */
set(MISSED) .
/* returns false */ .
. /* sees MISSED = 1 */
. /* so qdisc not empty */
. __qdisc_run()
. .
. pfifo_fast_dequeue()
----> /* may be done here */ .

|         .                                clear(MISSED)
|         .                                    .
|         .                                smp_mb __after_atomic();
|         .                                    .
|         .                                /* recheck the queue */
|         .                                /* nothing => exit   */
|   enqueue(skb1)
|         .
|   qdisc_run_begin()
|         .
|     spin_trylock() /* fail */
|         .
|     smp_mb__before_atomic() /* not enough */
|         .

---- if (test_bit(MISSED))
return false; /* exit */

In the above scenario, CPU 1 and CPU 2 both try to grab the qdisc->seqlock at the same time. Only CPU 2 succeeds and enters the bypass code path, where it emits its skb then calls __qdisc_run().

CPU1 fails, sets MISSED and goes down the traditionnal enqueue() + dequeue() code path. But when executing qdisc_run_begin() for the second time, after enqueuing its skbuff, it sees the MISSED bit still set (by itself) and consequently chooses to exit early without setting it again nor trying to grab the spinlock again.

Meanwhile CPU2 has seen MISSED = 1, cleared it, checked the queue and found it empty, so it returned.

At the end of the sequence, we end up with skb1 enqueued in the backlog, both CPUs out of __dev_xmit_skb(), the MISSED bit not set, and no __netif_schedule() called made. skb1 will now linger in the qdisc until somebody later performs a full __qdisc_run(). Associated to the bypass capacity of the qdisc, and the ability of the TCP layer to avoid resending packets which it knows are still in the qdisc, this can lead to serious traffic "holes" in a TCP connection.

We fix this by replacing the smp_mb__before_atomic() / test_bit() / set_bit() / smp_mb__after_atomic() sequence inside qdisc_run_begin() by a single test_and_set_bit() call, which is more concise and enforces the needed memory barriers.

Fixes 89837eb4b246 ("net: sched: add barrier to ensure correct ordering for lockless qdisc")
---
 include/net/sch_generic.h | 36 ++++++++----------------------------
 1 file changed, 8 insertions(+), 28 deletions(-)
include/net/sch_generic.h +8 -28
index 472843eedbae..4fe0892dde1a 100644
@@ -187,37 +187,17 @@ static inline bool qdisc_run_begin(struct Qdisc *qdisc)
 		if (spin_trylock(&qdisc->seqlock))
 			return true;
 
-		/* Paired with smp_mb__after_atomic() to make sure
-		 * STATE_MISSED checking is synchronized with clearing
-		 * in pfifo_fast_dequeue().
+		/* No need to insist if the MISSED flag was already set.
+		 * Note that test_and_set_bit() also gives us memory ordering
+		 * guarantees wrt potential earlier enqueue() and below
+		 * spin_trylock(), both of which are necessary to prevent races
 		 */
-		smp_mb__before_atomic();
-
-		/* If the MISSED flag is set, it means other thread has
-		 * set the MISSED flag before second spin_trylock(), so
-		 * we can return false here to avoid multi cpus doing
-		 * the set_bit() and second spin_trylock() concurrently.
-		 */
-		if (test_bit(__QDISC_STATE_MISSED, &qdisc->state))
+		if (test_and_set_bit(__QDISC_STATE_MISSED, &qdisc->state))
 			return false;
 
-		/* Set the MISSED flag before the second spin_trylock(),
-		 * if the second spin_trylock() return false, it means
-		 * other cpu holding the lock will do dequeuing for us
-		 * or it will see the MISSED flag set after releasing
-		 * lock and reschedule the net_tx_action() to do the
-		 * dequeuing.
-		 */
-		set_bit(__QDISC_STATE_MISSED, &qdisc->state);
-
-		/* spin_trylock() only has load-acquire semantic, so use
-		 * smp_mb__after_atomic() to ensure STATE_MISSED is set
-		 * before doing the second spin_trylock().
-		 */
-		smp_mb__after_atomic();
-
-		/* Retry again in case other CPU may not see the new flag
-		 * after it releases the lock at the end of qdisc_run_end().
+		/* Try to take the lock again to make sure that we will either
+		 * grab it or the CPU that still has it will see MISSED set
+		 * when testing it in qdisc_run_end()
 		 */
 		return spin_trylock(&qdisc->seqlock);
 	}
Signature · 1 line
2.35.1
author maintainer
#84 lore raw

PATCH 84/298 net: ethernet: mtk_eth_soc: out of bounds read in mtk_hwlro_get_fdir_entry()

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit e7e7104e2d5ddf3806a28695670f21bef471f1e1 ]

The "fsp->location" variable comes from user via ethtool_get_rxnfc(). Check that it is valid to prevent an out of bounds read.

Fixes 7aab747e5563 ("net: ethernet: mediatek: add ethtool functions to configure RX flows of HW LRO")
---
 drivers/net/ethernet/mediatek/mtk_eth_soc.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/net/ethernet/mediatek/mtk_eth_soc.c +3 -0
index f02d07ec5ccb..a50090e62c8f 100644
@@ -1949,6 +1949,9 @@ static int mtk_hwlro_get_fdir_entry(struct net_device *dev,
 	struct ethtool_rx_flow_spec *fsp =
 		(struct ethtool_rx_flow_spec *)&cmd->fs;
 
+	if (fsp->location >= ARRAY_SIZE(mac->hwlro_ip))
+		return -EINVAL;
+
 	/* only tcp dst ipv4 is meaningful, others are meaningless */
 	fsp->flow_type = TCP_V4_FLOW;
 	fsp->h_u.tcp_ip4_spec.ip4dst = ntohl(mac->hwlro_ip[fsp->location]);
Signature · 1 line
2.35.1
author maintainer
#85 lore raw

PATCH 85/298 net: ethernet: ti: am65-cpsw-nuss: Fix some refcount leaks

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 5dd89d2fc438457811cbbec07999ce0d80051ff5 ]

of_get_child_by_name() returns a node pointer with refcount incremented, we should use of_node_put() on it when not need anymore. am65_cpsw_init_cpts() and am65_cpsw_nuss_probe() don't release the refcount in error case.
Add missing of_node_put() to avoid refcount leak.

Fixes b1f66a5bee07 ("net: ethernet: ti: am65-cpsw-nuss: enable packet timestamping support")
Fixes 93a76530316a ("net: ethernet: ti: introduce am65x/j721e gigabit eth subsystem driver")
---
 drivers/net/ethernet/ti/am65-cpsw-nuss.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/ethernet/ti/am65-cpsw-nuss.c +2 -1
index 8251d7eb001b..eda91336c9f6 100644
@@ -1802,6 +1802,7 @@ static int am65_cpsw_init_cpts(struct am65_cpsw_common *common)
 	if (IS_ERR(cpts)) {
 		int ret = PTR_ERR(cpts);
 
+		of_node_put(node);
 		if (ret == -EOPNOTSUPP) {
 			dev_info(dev, "cpts disabled\n");
 			return 0;
@@ -2669,9 +2670,9 @@ static int am65_cpsw_nuss_probe(struct platform_device *pdev)
 	if (!node)
 		return -ENOENT;
 	common->port_num = of_get_child_count(node);
+	of_node_put(node);
 	if (common->port_num < 1 || common->port_num > AM65_CPSW_MAX_PORTS)
 		return -ENOENT;
-	of_node_put(node);
 
 	common->rx_flow_id_base = -1;
 	init_completion(&common->tdown_complete);
Signature · 1 line
2.35.1
author maintainer
#86 lore raw

PATCH 86/298 net: dsa: mv88e6xxx: Fix refcount leak in mv88e6xxx_mdios_register

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 02ded5a173619b11728b8bf75a3fd995a2c1ff28 ]

of_get_child_by_name() returns a node pointer with refcount incremented, we should use of_node_put() on it when done.

mv88e6xxx_mdio_register() pass the device node to of_mdiobus_register(). We don't need the device node after it.

Add missing of_node_put() to avoid refcount leak.

Fixes a3c53be55c95 ("net: dsa: mv88e6xxx: Support multiple MDIO busses")
---
 drivers/net/dsa/mv88e6xxx/chip.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/dsa/mv88e6xxx/chip.c +1 -0
index cf7754dddad7..283ae376f469 100644
@@ -3482,6 +3482,7 @@ static int mv88e6xxx_mdios_register(struct mv88e6xxx_chip *chip,
 	 */
 	child = of_get_child_by_name(np, "mdio");
 	err = mv88e6xxx_mdio_register(chip, child, false);
+	of_node_put(child);
 	if (err)
 		return err;
 
Signature · 1 line
2.35.1
author maintainer
#87 lore raw

PATCH 87/298 modpost: fix removing numeric suffixes

In reply to Greg Kroah-Hartman

From: Alexander Lobakin <[email protected]>

[ Upstream commit b5beffa20d83c4e15306c991ffd00de0d8628338 ]

With the -z unique-symbol linker flag or any similar mechanism, it is possible to trigger the following:

ERROR: modpost: "param_set_uint.0" [vmlinux] is a static EXPORT_SYMBOL

The reason is that for now the condition from remove_dot():

if (m && (s[n + m] == '.' || s[n + m] == 0))

which was designed to test if it's a dot or a '\0' after the suffix is never satisfied.
This is due to that s[n + m] always points to the last digit of a numeric suffix, not on the symbol next to it (from a custom debug print added to modpost):

param_set_uint.0, s[n + m] is '0', s[n + m + 1] is '\0'

So it's off-by-one and was like that since 2014.

Fix this for the sake of any potential upcoming features, but don't bother stable-backporting, as it's well hidden -- apart from that LD flag, it can be triggered only with GCC LTO which never landed upstream.

Fixes fcd38ed0ff26 ("scripts: modpost: fix compilation warning")
---
 scripts/mod/modpost.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
scripts/mod/modpost.c +1 -1
index e04ae56931e2..3b098e26938e 100644
@@ -1986,7 +1986,7 @@ static char *remove_dot(char *s)
 
 	if (n && s[n]) {
 		size_t m = strspn(s + n + 1, "0123456789");
-		if (m && (s[n + m] == '.' || s[n + m] == 0))
+		if (m && (s[n + m + 1] == '.' || s[n + m + 1] == 0))
 			s[n] = 0;
 
 		/* strip trailing .lto */
Signature · 1 line
2.35.1
author maintainer
#88 lore raw

PATCH 88/298 ep93xx: clock: Do not return the address of the freed memory

In reply to Greg Kroah-Hartman

From: Genjian Zhang <[email protected]>

[ Upstream commit 8a7322a3a05f75e8a4902bdf8129aecd37d54fe9 ]

Avoid return freed memory addresses,Modified to the actual error return value of clk_register().

Fixes 9645ccc7bd7a ("ep93xx: clock: convert in-place to COMMON_CLK")
---
 arch/arm/mach-ep93xx/clock.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
arch/arm/mach-ep93xx/clock.c +6 -4
index 28e0ae6e890e..00c2db101ce5 100644
@@ -345,9 +345,10 @@ static struct clk_hw *clk_hw_register_ddiv(const char *name,
 	psc->hw.init = &init;
 
 	clk = clk_register(NULL, &psc->hw);
-	if (IS_ERR(clk))
+	if (IS_ERR(clk)) {
 		kfree(psc);
-
+		return ERR_CAST(clk);
+	}
 	return &psc->hw;
 }
 
@@ -452,9 +453,10 @@ static struct clk_hw *clk_hw_register_div(const char *name,
 	psc->hw.init = &init;
 
 	clk = clk_register(NULL, &psc->hw);
-	if (IS_ERR(clk))
+	if (IS_ERR(clk)) {
 		kfree(psc);
-
+		return ERR_CAST(clk);
+	}
 	return &psc->hw;
 }
 
Signature · 1 line
2.35.1
author maintainer
#89 lore raw

PATCH 89/298 jffs2: fix memory leak in jffs2_do_fill_super

In reply to Greg Kroah-Hartman

From: Baokun Li <[email protected]>

[ Upstream commit c14adb1cf70a984ed081c67e9d27bc3caad9537c ]

If jffs2_iget() or d_make_root() in jffs2_do_fill_super() returns an error, we can observe the following kmemleak report:

--------------------------------------------
unreferenced object 0xffff888105a65340 (size 64):
comm "mount", pid 710, jiffies 4302851558 (age 58.239s) hex dump (first 32 bytes):

    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff859c45e5>] kmem_cache_alloc_trace+0x475/0x8a0
    [<ffffffff86160146>] jffs2_sum_init+0x96/0x1a0
    [<ffffffff86140e25>] jffs2_do_mount_fs+0x745/0x2120
    [<ffffffff86149fec>] jffs2_do_fill_super+0x35c/0x810
    [<ffffffff8614aae9>] jffs2_fill_super+0x2b9/0x3b0
    [...]

unreferenced object 0xffff8881bd7f0000 (size 65536):
comm "mount", pid 710, jiffies 4302851558 (age 58.239s) hex dump (first 32 bytes):
bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb ................ bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb ................
backtrace:

    [<ffffffff858579ba>] kmalloc_order+0xda/0x110
    [<ffffffff85857a11>] kmalloc_order_trace+0x21/0x130
    [<ffffffff859c2ed1>] __kmalloc+0x711/0x8a0
    [<ffffffff86160189>] jffs2_sum_init+0xd9/0x1a0
    [<ffffffff86140e25>] jffs2_do_mount_fs+0x745/0x2120
    [<ffffffff86149fec>] jffs2_do_fill_super+0x35c/0x810
    [<ffffffff8614aae9>] jffs2_fill_super+0x2b9/0x3b0
    [...]
--------------------------------------------

This is because the resources allocated in jffs2_sum_init() are not released. Call jffs2_sum_exit() to release these resources to solve the problem.

Fixes e631ddba5887 ("[JFFS2] Add erase block summary support (mount time improvement)")
---
 fs/jffs2/fs.c | 1 +
 1 file changed, 1 insertion(+)
fs/jffs2/fs.c +1 -0
index 71f03a5d36ed..f83a468b6488 100644
@@ -604,6 +604,7 @@ int jffs2_do_fill_super(struct super_block *sb, struct fs_context *fc)
 	jffs2_free_raw_node_refs(c);
 	kvfree(c->blocks);
 	jffs2_clear_xattr_subsystem(c);
+	jffs2_sum_exit(c);
  out_inohash:
 	kfree(c->inocache_list);
  out_wbuf:
Signature · 1 line
2.35.1
author maintainer
#90 lore raw

PATCH 90/298 ubi: fastmap: Fix high cpu usage of ubi_bgt by making sure wl_pool not empty

In reply to Greg Kroah-Hartman

From: Zhihao Cheng <[email protected]>

[ Upstream commit d09e9a2bddba6c48e0fddb16c4383172ac593251 ]

There at least 6 PEBs reserved on UBI device:
1. EBA_RESERVED_PEBS[1]
2. WL_RESERVED_PEBS[1]
3. UBI_LAYOUT_VOLUME_EBS[2]
4. MIN_FASTMAP_RESERVED_PEBS[2]

When all ubi volumes take all their PEBs, there are 3 (EBA_RESERVED_PEBS + WL_RESERVED_PEBS + MIN_FASTMAP_RESERVED_PEBS - MIN_FASTMAP_TAKEN_PEBS[1]) free PEBs. Since commit f9c34bb529975fe ("ubi: Fix producing anchor PEBs") and commit 4b68bf9a69d22dd ("ubi: Select fastmap anchor PEBs considering wear level rules") applied, there is only 1 (3 - FASTMAP_ANCHOR_PEBS[1] - FASTMAP_NEXT_ANCHOR_PEBS[1]) free PEB to fill pool and wl_pool, after filling pool, wl_pool is always empty. So, UBI could be stuck in an infinite loop:

ubi_thread system_wq
wear_leveling_worker <--------------------------------------------------
get_peb_for_wl |

    // fm_wl_pool, used = size = 0					|
    schedule_work(&ubi->fm_work)					|
									|
		    update_fastmap_work_fn				|
		      ubi_update_fastmap				|
			ubi_refill_pools				|
			// ubi->free_count - ubi->beb_rsvd_pebs < 5	|
			// wl_pool is not filled with any PEBs		|
			schedule_erase(old_fm_anchor)			|
			ubi_ensure_anchor_pebs				|
			  __schedule_ubi_work(wear_leveling_worker)	|
									|
__erase_worker								|
  ensure_wear_leveling							|
    __schedule_ubi_work(wear_leveling_worker) --------------------------

, which cause high cpu usage of ubi_bgt:
top - 12:10:42 up 5 min, 2 users, load average: 1.76, 0.68, 0.27
Tasks: 123 total, 3 running, 54 sleeping, 0 stopped, 0 zombie

  PID USER PR   NI VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
 1589 root 20   0   0      0      0 R  45.0  0.0   0:38.86 ubi_bgt0d
  319 root 20   0   0      0      0 I  15.2  0.0   0:15.29 kworker/0:3-eve
  371 root 20   0   0      0      0 I  14.9  0.0   0:12.85 kworker/3:3-eve
   20 root 20   0   0      0      0 I  11.3  0.0   0:05.33 kworker/1:0-eve
  202 root 20   0   0      0      0 I  11.3  0.0   0:04.93 kworker/2:3-eve

In commit 4b68bf9a69d22dd ("ubi: Select fastmap anchor PEBs considering wear level rules"), there are three key changes:
1) Choose the fastmap anchor when the most free PEBs are available.
2) Enable anchor move within the anchor area again as it is useful for distributing wear.
3) Import a candidate fm anchor and check this PEB's erase count during wear leveling. If the wear leveling limit is exceeded, use the used anchor area PEB with the lowest erase count to replace it.

The anchor candidate can be removed, we can check fm_anchor PEB's erase count during wear leveling. Fix it by:
1) Removing 'fm_next_anchor' and check 'fm_anchor' during wear leveling.
2) Preferentially filling one free peb into fm_wl_pool in condition of ubi->free_count > ubi->beb_rsvd_pebs, then try to reserve enough free count for fastmap non anchor pebs after the above prerequisites are met.
Then, there are at least 1 PEB in pool and 1 PEB in wl_pool after calling ubi_refill_pools() with all erase works done.

Fetch a reproducer in [Link].

Fixes 4b68bf9a69d22dd ("ubi: Select fastmap anchor PEBs ... rules")
---
 drivers/mtd/ubi/fastmap-wl.c | 69 ++++++++++++++++++++++++------------
 drivers/mtd/ubi/fastmap.c    | 11 ------
 drivers/mtd/ubi/ubi.h        |  4 +--
 drivers/mtd/ubi/wl.c         | 19 +++++-----
 4 files changed, 57 insertions(+), 46 deletions(-)
drivers/mtd/ubi/fastmap-wl.c +46 -23
index 28f55f9cf715..053ab52668e8 100644
@@ -97,6 +97,33 @@ struct ubi_wl_entry *ubi_wl_get_fm_peb(struct ubi_device *ubi, int anchor)
 	return e;
 }
 
+/*
+ * has_enough_free_count - whether ubi has enough free pebs to fill fm pools
+ * @ubi: UBI device description object
+ * @is_wl_pool: whether UBI is filling wear leveling pool
+ *
+ * This helper function checks whether there are enough free pebs (deducted
+ * by fastmap pebs) to fill fm_pool and fm_wl_pool, above rule works after
+ * there is at least one of free pebs is filled into fm_wl_pool.
+ * For wear leveling pool, UBI should also reserve free pebs for bad pebs
+ * handling, because there maybe no enough free pebs for user volumes after
+ * producing new bad pebs.
+ */
+static bool has_enough_free_count(struct ubi_device *ubi, bool is_wl_pool)
+{
+	int fm_used = 0;	// fastmap non anchor pebs.
+	int beb_rsvd_pebs;
+
+	if (!ubi->free.rb_node)
+		return false;
+
+	beb_rsvd_pebs = is_wl_pool ? ubi->beb_rsvd_pebs : 0;
+	if (ubi->fm_wl_pool.size > 0 && !(ubi->ro_mode || ubi->fm_disabled))
+		fm_used = ubi->fm_size / ubi->leb_size - 1;
+
+	return ubi->free_count - beb_rsvd_pebs > fm_used;
+}
+
 /**
  * ubi_refill_pools - refills all fastmap PEB pools.
  * @ubi: UBI device description object
@@ -120,21 +147,17 @@ void ubi_refill_pools(struct ubi_device *ubi)
 		wl_tree_add(ubi->fm_anchor, &ubi->free);
 		ubi->free_count++;
 	}
-	if (ubi->fm_next_anchor) {
-		wl_tree_add(ubi->fm_next_anchor, &ubi->free);
-		ubi->free_count++;
-	}
 
-	/* All available PEBs are in ubi->free, now is the time to get
+	/*
+	 * All available PEBs are in ubi->free, now is the time to get
 	 * the best anchor PEBs.
 	 */
 	ubi->fm_anchor = ubi_wl_get_fm_peb(ubi, 1);
-	ubi->fm_next_anchor = ubi_wl_get_fm_peb(ubi, 1);
 
 	for (;;) {
 		enough = 0;
 		if (pool->size < pool->max_size) {
-			if (!ubi->free.rb_node)
+			if (!has_enough_free_count(ubi, false))
 				break;
 
 			e = wl_get_wle(ubi);
@@ -147,8 +170,7 @@ void ubi_refill_pools(struct ubi_device *ubi)
 			enough++;
 
 		if (wl_pool->size < wl_pool->max_size) {
-			if (!ubi->free.rb_node ||
-			   (ubi->free_count - ubi->beb_rsvd_pebs < 5))
+			if (!has_enough_free_count(ubi, true))
 				break;
 
 			e = find_wl_entry(ubi, &ubi->free, WL_FREE_MAX_DIFF);
@@ -286,20 +308,26 @@ static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi)
 int ubi_ensure_anchor_pebs(struct ubi_device *ubi)
 {
 	struct ubi_work *wrk;
+	struct ubi_wl_entry *anchor;
 
 	spin_lock(&ubi->wl_lock);
 
-	/* Do we have a next anchor? */
-	if (!ubi->fm_next_anchor) {
-		ubi->fm_next_anchor = ubi_wl_get_fm_peb(ubi, 1);
-		if (!ubi->fm_next_anchor)
-			/* Tell wear leveling to produce a new anchor PEB */
-			ubi->fm_do_produce_anchor = 1;
+	/* Do we already have an anchor? */
+	if (ubi->fm_anchor) {
+		spin_unlock(&ubi->wl_lock);
+		return 0;
 	}
 
-	/* Do wear leveling to get a new anchor PEB or check the
-	 * existing next anchor candidate.
-	 */
+	/* See if we can find an anchor PEB on the list of free PEBs */
+	anchor = ubi_wl_get_fm_peb(ubi, 1);
+	if (anchor) {
+		ubi->fm_anchor = anchor;
+		spin_unlock(&ubi->wl_lock);
+		return 0;
+	}
+
+	ubi->fm_do_produce_anchor = 1;
+	/* No luck, trigger wear leveling to produce a new anchor PEB. */
 	if (ubi->wl_scheduled) {
 		spin_unlock(&ubi->wl_lock);
 		return 0;
@@ -381,11 +409,6 @@ static void ubi_fastmap_close(struct ubi_device *ubi)
 		ubi->fm_anchor = NULL;
 	}
 
-	if (ubi->fm_next_anchor) {
-		return_unused_peb(ubi, ubi->fm_next_anchor);
-		ubi->fm_next_anchor = NULL;
-	}
-
 	if (ubi->fm) {
 		for (i = 0; i < ubi->fm->used_blocks; i++)
 			kfree(ubi->fm->e[i]);
drivers/mtd/ubi/fastmap.c +0 -11
index 6b5f1ffd961b..6e95c4b1473e 100644
@@ -1230,17 +1230,6 @@ static int ubi_write_fastmap(struct ubi_device *ubi,
 		fm_pos += sizeof(*fec);
 		ubi_assert(fm_pos <= ubi->fm_size);
 	}
-	if (ubi->fm_next_anchor) {
-		fec = (struct ubi_fm_ec *)(fm_raw + fm_pos);
-
-		fec->pnum = cpu_to_be32(ubi->fm_next_anchor->pnum);
-		set_seen(ubi, ubi->fm_next_anchor->pnum, seen_pebs);
-		fec->ec = cpu_to_be32(ubi->fm_next_anchor->ec);
-
-		free_peb_count++;
-		fm_pos += sizeof(*fec);
-		ubi_assert(fm_pos <= ubi->fm_size);
-	}
 	fmh->free_peb_count = cpu_to_be32(free_peb_count);
 
 	ubi_for_each_used_peb(ubi, wl_e, tmp_rb) {
drivers/mtd/ubi/ubi.h +1 -3
index 7c083ad58274..078112e23dfd 100644
@@ -489,8 +489,7 @@ struct ubi_debug_info {
  * @fm_work: fastmap work queue
  * @fm_work_scheduled: non-zero if fastmap work was scheduled
  * @fast_attach: non-zero if UBI was attached by fastmap
- * @fm_anchor: The new anchor PEB used during fastmap update
- * @fm_next_anchor: An anchor PEB candidate for the next time fastmap is updated
+ * @fm_anchor: The next anchor PEB to use for fastmap
  * @fm_do_produce_anchor: If true produce an anchor PEB in wl
  *
  * @used: RB-tree of used physical eraseblocks
@@ -601,7 +600,6 @@ struct ubi_device {
 	int fm_work_scheduled;
 	int fast_attach;
 	struct ubi_wl_entry *fm_anchor;
-	struct ubi_wl_entry *fm_next_anchor;
 	int fm_do_produce_anchor;
 
 	/* Wear-leveling sub-system's stuff */
drivers/mtd/ubi/wl.c +10 -9
index 8455f1d47f3c..afcdacb9d0e9 100644
@@ -689,16 +689,16 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk,
 
 #ifdef CONFIG_MTD_UBI_FASTMAP
 	e1 = find_anchor_wl_entry(&ubi->used);
-	if (e1 && ubi->fm_next_anchor &&
-	    (ubi->fm_next_anchor->ec - e1->ec >= UBI_WL_THRESHOLD)) {
+	if (e1 && ubi->fm_anchor &&
+	    (ubi->fm_anchor->ec - e1->ec >= UBI_WL_THRESHOLD)) {
 		ubi->fm_do_produce_anchor = 1;
-		/* fm_next_anchor is no longer considered a good anchor
-		 * candidate.
+		/*
+		 * fm_anchor is no longer considered a good anchor.
 		 * NULL assignment also prevents multiple wear level checks
 		 * of this PEB.
 		 */
-		wl_tree_add(ubi->fm_next_anchor, &ubi->free);
-		ubi->fm_next_anchor = NULL;
+		wl_tree_add(ubi->fm_anchor, &ubi->free);
+		ubi->fm_anchor = NULL;
 		ubi->free_count++;
 	}
 
@@ -1085,12 +1085,13 @@ static int __erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk)
 	if (!err) {
 		spin_lock(&ubi->wl_lock);
 
-		if (!ubi->fm_disabled && !ubi->fm_next_anchor &&
+		if (!ubi->fm_disabled && !ubi->fm_anchor &&
 		    e->pnum < UBI_FM_MAX_START) {
-			/* Abort anchor production, if needed it will be
+			/*
+			 * Abort anchor production, if needed it will be
 			 * enabled again in the wear leveling started below.
 			 */
-			ubi->fm_next_anchor = e;
+			ubi->fm_anchor = e;
 			ubi->fm_do_produce_anchor = 0;
 		} else {
 			wl_tree_add(e, &ubi->free);
Signature · 1 line
2.35.1
author maintainer
#91 lore raw

PATCH 91/298 ubi: ubi_create_volume: Fix use-after-free when volume creation failed

In reply to Greg Kroah-Hartman

From: Zhihao Cheng <[email protected]>

[ Upstream commit 8c03a1c21d72210f81cb369cc528e3fde4b45411 ]

There is an use-after-free problem for 'eba_tbl' in ubi_create_volume()'s error handling path:

ubi_eba_replace_table(vol, eba_tbl)

    vol->eba_tbl = tbl

out_mapping:
ubi_eba_destroy_table(eba_tbl) // Free 'eba_tbl'
out_unlock:
put_device(&vol->dev)

    vol_release
      kfree(tbl->entries)	  // UAF

Fix it by removing redundant 'eba_tbl' releasing. Fetch a reproducer in [Link].

Fixes 493cfaeaa0c9b ("mtd: utilize new cdev_device_add helper function")
---
 drivers/mtd/ubi/vmt.c | 1 -
 1 file changed, 1 deletion(-)
drivers/mtd/ubi/vmt.c +0 -1
index 1bc7b3a05604..6ea95ade4ca6 100644
@@ -309,7 +309,6 @@ int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req)
 	ubi->volumes[vol_id] = NULL;
 	ubi->vol_count -= 1;
 	spin_unlock(&ubi->volumes_lock);
-	ubi_eba_destroy_table(eba_tbl);
 out_acc:
 	spin_lock(&ubi->volumes_lock);
 	ubi->rsvd_pebs -= vol->reserved_pebs;
Signature · 1 line
2.35.1
author maintainer
#92 lore raw

PATCH 92/298 selftests/bpf: fix selftest after random: Urandom_read tracepoint removal

In reply to Greg Kroah-Hartman

From: Andrii Nakryiko <[email protected]>

[ Upstream commit 99dea2c664d7bc7e4f6f6947182d0d365165a998 ]

14c174633f34 ("random: remove unused tracepoints") removed all the tracepoints from drivers/char/random.c, one of which, random:urandom_read, was used by stacktrace_build_id selftest to trigger stack trace capture.

Fix breakage by switching to kprobing urandom_read() function.

---
 .../selftests/bpf/progs/test_stacktrace_build_id.c   | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)
tools/testing/selftests/bpf/progs/test_stacktrace_build_id.c +2 -10
index 36a707e7c7a7..6c62bfb8bb6f 100644
@@ -39,16 +39,8 @@ struct {
 	__type(value, stack_trace_t);
 } stack_amap SEC(".maps");
 
-/* taken from /sys/kernel/debug/tracing/events/random/urandom_read/format */
-struct random_urandom_args {
-	unsigned long long pad;
-	int got_bits;
-	int pool_left;
-	int input_left;
-};
-
-SEC("tracepoint/random/urandom_read")
-int oncpu(struct random_urandom_args *args)
+SEC("kprobe/urandom_read")
+int oncpu(struct pt_regs *args)
 {
 	__u32 max_len = sizeof(struct bpf_stack_build_id)
 			* PERF_MAX_STACK_DEPTH;
Signature · 1 line
2.35.1
author maintainer
#93 lore raw

PATCH 93/298 selftests/bpf: fix stacktrace_build_id with missing kprobe/urandom_read

In reply to Greg Kroah-Hartman

From: Song Liu <[email protected]>

[ Upstream commit 59ed76fe2f981bccde37bdddb465f260a96a2404 ]

Kernel function urandom_read is replaced with urandom_read_iter. Therefore, kprobe on urandom_read is not working any more:

[root@eth50-1 bpf]# ./test_progs -n 161 test_stacktrace_build_id:PASS:skel_open_and_load 0 nsec
libbpf: kprobe perf_event_open() failed: No such file or directory
libbpf: prog 'oncpu': failed to create kprobe 'urandom_read+0x0' \
perf event: No such file or directory
libbpf: prog 'oncpu': failed to auto-attach: -2 test_stacktrace_build_id:FAIL:attach_tp err -2
161 stacktrace_build_id:FAIL

Fix this by replacing urandom_read with urandom_read_iter in the test.

---
 tools/testing/selftests/bpf/progs/test_stacktrace_build_id.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/testing/selftests/bpf/progs/test_stacktrace_build_id.c +1 -1
index 6c62bfb8bb6f..0c4426592a26 100644
@@ -39,7 +39,7 @@ struct {
 	__type(value, stack_trace_t);
 } stack_amap SEC(".maps");
 
-SEC("kprobe/urandom_read")
+SEC("kprobe/urandom_read_iter")
 int oncpu(struct pt_regs *args)
 {
 	__u32 max_len = sizeof(struct bpf_stack_build_id)
Signature · 1 line
2.35.1
author maintainer
#94 lore raw

PATCH 94/298 bpf: Fix probe read error in ___bpf_prog_run()

In reply to Greg Kroah-Hartman

From: Menglong Dong <[email protected]>

[ Upstream commit caff1fa4118cec4dfd4336521ebd22a6408a1e3e ]

I think there is something wrong with BPF_PROBE_MEM in ___bpf_prog_run() in big-endian machine. Let's make a test and see what will happen if we want to load a 'u16' with BPF_PROBE_MEM.

Let's make the src value '0x0001', the value of dest register will become 0x0001000000000000, as the value will be loaded to the first 2 byte of DST with following code:

  bpf_probe_read_kernel(&DST, SIZE, (const void *)(long) (SRC + insn->off));

Obviously, the value in DST is not correct. In fact, we can compare BPF_PROBE_MEM with LDX_MEM_H:

  DST = *(SIZE *)(unsigned long) (SRC + insn->off);

If the memory load is done by LDX_MEM_H, the value in DST will be 0x1 now.

And I think this error results in the test case 'test_bpf_sk_storage_map' failing:

test_bpf_sk_storage_map:PASS:bpf_iter_bpf_sk_storage_map__open_and_load 0 nsec test_bpf_sk_storage_map:PASS:socket 0 nsec test_bpf_sk_storage_map:PASS:map_update 0 nsec test_bpf_sk_storage_map:PASS:socket 0 nsec test_bpf_sk_storage_map:PASS:map_update 0 nsec test_bpf_sk_storage_map:PASS:socket 0 nsec test_bpf_sk_storage_map:PASS:map_update 0 nsec test_bpf_sk_storage_map:PASS:attach_iter 0 nsec test_bpf_sk_storage_map:PASS:create_iter 0 nsec test_bpf_sk_storage_map:PASS:read 0 nsec test_bpf_sk_storage_map:FAIL:ipv6_sk_count got 0 expected 3 $10/26 bpf_iter/bpf_sk_storage_map:FAIL

The code of the test case is simply, it will load sk->sk_family to the register with BPF_PROBE_MEM and check if it is AF_INET6. With this patch, now the test case 'bpf_iter' can pass:

$10 bpf_iter:OK

Fixes 2a02759ef5f8 ("bpf: Add support for BTF pointers to interpreter")
Cc Ilya Leoshkevich <[email protected]>
---
 kernel/bpf/core.c | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)
kernel/bpf/core.c +5 -9
index 64c44eed8c07..10c4a2028e07 100644
@@ -1671,6 +1671,11 @@ static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn)
 		CONT;							\
 	LDX_MEM_##SIZEOP:						\
 		DST = *(SIZE *)(unsigned long) (SRC + insn->off);	\
+		CONT;							\
+	LDX_PROBE_MEM_##SIZEOP:						\
+		bpf_probe_read_kernel(&DST, sizeof(SIZE),		\
+				      (const void *)(long) (SRC + insn->off));	\
+		DST = *((SIZE *)&DST);					\
 		CONT;
 
 	LDST(B,   u8)
@@ -1678,15 +1683,6 @@ static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn)
 	LDST(W,  u32)
 	LDST(DW, u64)
 #undef LDST
-#define LDX_PROBE(SIZEOP, SIZE)							\
-	LDX_PROBE_MEM_##SIZEOP:							\
-		bpf_probe_read_kernel(&DST, SIZE, (const void *)(long) (SRC + insn->off));	\
-		CONT;
-	LDX_PROBE(B,  1)
-	LDX_PROBE(H,  2)
-	LDX_PROBE(W,  4)
-	LDX_PROBE(DW, 8)
-#undef LDX_PROBE
 
 #define ATOMIC_ALU_OP(BOP, KOP)						\
 		case BOP:						\
Signature · 1 line
2.35.1
author maintainer
#95 lore raw

PATCH 95/298 block: take destination bvec offsets into account in bio_copy_data_iter

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit 403d50341cce6b5481a92eb481e6df60b1f49b55 ]

Appartly bcache can copy into bios that do not just contain fresh pages but can have offsets into the bio_vecs. Restore support for tht in bio_copy_data_iter.

Fixes f8b679a070c5 ("block: rewrite bio_copy_data_iter to use bvec_kmap_local and memcpy_to_bvec")
---
 block/bio.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
block/bio.c +5 -3
index 342b1cf5d713..738fea03edbf 100644
@@ -1308,10 +1308,12 @@ void bio_copy_data_iter(struct bio *dst, struct bvec_iter *dst_iter,
 		struct bio_vec src_bv = bio_iter_iovec(src, *src_iter);
 		struct bio_vec dst_bv = bio_iter_iovec(dst, *dst_iter);
 		unsigned int bytes = min(src_bv.bv_len, dst_bv.bv_len);
-		void *src_buf;
+		void *src_buf = bvec_kmap_local(&src_bv);
+		void *dst_buf = bvec_kmap_local(&dst_bv);
 
-		src_buf = bvec_kmap_local(&src_bv);
-		memcpy_to_bvec(&dst_bv, src_buf);
+		memcpy(dst_buf, src_buf, bytes);
+
+		kunmap_local(dst_buf);
 		kunmap_local(src_buf);
 
 		bio_advance_iter_single(src, src_iter, bytes);
Signature · 1 line
2.35.1
author maintainer
#96 lore raw

PATCH 96/298 nbd: dont clear NBD_CMD_INFLIGHT flag if request is not completed

In reply to Greg Kroah-Hartman

From: Yu Kuai <[email protected]>

[ Upstream commit 2895f1831e911ca87d4efdf43e35eb72a0c7e66e ]

Otherwise io will hung because request will only be completed if the cmd has the flag 'NBD_CMD_INFLIGHT'.

Fixes 07175cb1baf4 ("nbd: make sure request completion won't concurrent")
---
 drivers/block/nbd.c | 18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)
drivers/block/nbd.c +14 -4
index 284557041336..ed678037ba6d 100644
@@ -404,13 +404,14 @@ static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req,
 	if (!mutex_trylock(&cmd->lock))
 		return BLK_EH_RESET_TIMER;
 
-	if (!__test_and_clear_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
+	if (!test_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
 		mutex_unlock(&cmd->lock);
 		return BLK_EH_DONE;
 	}
 
 	if (!refcount_inc_not_zero(&nbd->config_refs)) {
 		cmd->status = BLK_STS_TIMEOUT;
+		__clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
 		mutex_unlock(&cmd->lock);
 		goto done;
 	}
@@ -479,6 +480,7 @@ static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req,
 	dev_err_ratelimited(nbd_to_dev(nbd), "Connection timed out\n");
 	set_bit(NBD_RT_TIMEDOUT, &config->runtime_flags);
 	cmd->status = BLK_STS_IOERR;
+	__clear_bit(NBD_CMD_INFLIGHT, &cmd->flags);
 	mutex_unlock(&cmd->lock);
 	sock_shutdown(nbd);
 	nbd_config_put(nbd);
@@ -746,7 +748,7 @@ static struct nbd_cmd *nbd_handle_reply(struct nbd_device *nbd, int index,
 	cmd = blk_mq_rq_to_pdu(req);
 
 	mutex_lock(&cmd->lock);
-	if (!__test_and_clear_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
+	if (!test_bit(NBD_CMD_INFLIGHT, &cmd->flags)) {
 		dev_err(disk_to_dev(nbd->disk), "Suspicious reply %d (status %u flags %lu)",
 			tag, cmd->status, cmd->flags);
 		ret = -ENOENT;
@@ -855,8 +857,16 @@ static void recv_work(struct work_struct *work)
 		}
 
 		rq = blk_mq_rq_from_pdu(cmd);
-		if (likely(!blk_should_fake_timeout(rq->q)))
-			blk_mq_complete_request(rq);
+		if (likely(!blk_should_fake_timeout(rq->q))) {
+			bool complete;
+
+			mutex_lock(&cmd->lock);
+			complete = __test_and_clear_bit(NBD_CMD_INFLIGHT,
+							&cmd->flags);
+			mutex_unlock(&cmd->lock);
+			if (complete)
+				blk_mq_complete_request(rq);
+		}
 		percpu_ref_put(&q->q_usage_counter);
 	}
 
Signature · 1 line
2.35.1
author maintainer
#97 lore raw

PATCH 97/298 riscv: read-only pages should not be writable

In reply to Greg Kroah-Hartman

From: Heinrich Schuchardt <[email protected]>

[ Upstream commit 630f972d76d6460235e84e1aa034ee06f9c8c3a9 ]

If EFI pages are marked as read-only,
we should remove the _PAGE_WRITE flag.

The current code overwrites an unused value.

---
 arch/riscv/kernel/efi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/riscv/kernel/efi.c +1 -1
index 024159298231..1aa540350abd 100644
@@ -65,7 +65,7 @@ static int __init set_permissions(pte_t *ptep, unsigned long addr, void *data)
 
 	if (md->attribute & EFI_MEMORY_RO) {
 		val = pte_val(pte) & ~_PAGE_WRITE;
-		val = pte_val(pte) | _PAGE_READ;
+		val |= _PAGE_READ;
 		pte = __pte(val);
 	}
 	if (md->attribute & EFI_MEMORY_XP) {
Signature · 1 line
2.35.1
author maintainer
#98 lore raw

PATCH 98/298 net/smc: fixes for converting from "struct smc_cdc_tx_pend **" to "struct smc_wr_tx_pend_priv *"

In reply to Greg Kroah-Hartman

From: Guangguan Wang <[email protected]>

[ Upstream commit e225c9a5a74b12e9ef8516f30a3db2c7eb866ee1 ]

"struct smc_cdc_tx_pend **" can not directly convert to "struct smc_wr_tx_pend_priv *".

Fixes 2bced6aefa3d ("net/smc: put slot when connection is killed")
---
 net/smc/smc_cdc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/smc/smc_cdc.c +1 -1
index 9d5a97168969..93042ef6869b 100644
@@ -72,7 +72,7 @@ int smc_cdc_get_free_slot(struct smc_connection *conn,
 		/* abnormal termination */
 		if (!rc)
 			smc_wr_tx_put_slot(link,
-					   (struct smc_wr_tx_pend_priv *)pend);
+					   (struct smc_wr_tx_pend_priv *)(*pend));
 		rc = -EPIPE;
 	}
 	return rc;
Signature · 1 line
2.35.1
author maintainer
#99 lore raw

PATCH 99/298 tcp: add accessors to read/set tp->snd_cwnd

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit 40570375356c874b1578e05c1dcc3ff7c1322dbe ]

We had various bugs over the years with code breaking the assumption that tp->snd_cwnd is greater than zero.

Lately, syzbot reported the WARN_ON_ONCE(!tp->prior_cwnd) added in commit 8b8a321ff72c ("tcp: fix zero cwnd in tcp_cwnd_reduction") can trigger, and without a repro we would have to spend considerable time finding the bug.

Instead of complaining too late, we want to catch where and when tp->snd_cwnd is set to an illegal value.

---
 include/net/tcp.h          | 19 +++++++++++++++----
 include/trace/events/tcp.h |  2 +-
 net/core/filter.c          |  2 +-
 net/ipv4/tcp.c             |  8 ++++----
 net/ipv4/tcp_bbr.c         | 20 ++++++++++----------
 net/ipv4/tcp_bic.c         | 14 +++++++-------
 net/ipv4/tcp_cdg.c         | 30 +++++++++++++++---------------
 net/ipv4/tcp_cong.c        | 18 +++++++++---------
 net/ipv4/tcp_cubic.c       | 22 +++++++++++-----------
 net/ipv4/tcp_dctcp.c       | 11 ++++++-----
 net/ipv4/tcp_highspeed.c   | 18 +++++++++---------
 net/ipv4/tcp_htcp.c        | 10 +++++-----
 net/ipv4/tcp_hybla.c       | 18 +++++++++---------
 net/ipv4/tcp_illinois.c    | 12 +++++++-----
 net/ipv4/tcp_input.c       | 36 ++++++++++++++++++------------------
 net/ipv4/tcp_ipv4.c        |  2 +-
 net/ipv4/tcp_lp.c          |  6 +++---
 net/ipv4/tcp_metrics.c     | 12 ++++++------
 net/ipv4/tcp_nv.c          | 24 ++++++++++++------------
 net/ipv4/tcp_output.c      | 30 +++++++++++++++---------------
 net/ipv4/tcp_rate.c        |  2 +-
 net/ipv4/tcp_scalable.c    |  4 ++--
 net/ipv4/tcp_vegas.c       | 21 +++++++++++----------
 net/ipv4/tcp_veno.c        | 24 ++++++++++++------------
 net/ipv4/tcp_westwood.c    |  3 ++-
 net/ipv4/tcp_yeah.c        | 30 +++++++++++++++---------------
 net/ipv6/tcp_ipv6.c        |  2 +-
 27 files changed, 208 insertions(+), 192 deletions(-)
include/net/tcp.h +15 -4
index a3fe2f9bc01c..818ac8077381 100644
@@ -1215,9 +1215,20 @@ static inline unsigned int tcp_packets_in_flight(const struct tcp_sock *tp)
 
 #define TCP_INFINITE_SSTHRESH	0x7fffffff
 
+static inline u32 tcp_snd_cwnd(const struct tcp_sock *tp)
+{
+	return tp->snd_cwnd;
+}
+
+static inline void tcp_snd_cwnd_set(struct tcp_sock *tp, u32 val)
+{
+	WARN_ON_ONCE((int)val <= 0);
+	tp->snd_cwnd = val;
+}
+
 static inline bool tcp_in_slow_start(const struct tcp_sock *tp)
 {
-	return tp->snd_cwnd < tp->snd_ssthresh;
+	return tcp_snd_cwnd(tp) < tp->snd_ssthresh;
 }
 
 static inline bool tcp_in_initial_slowstart(const struct tcp_sock *tp)
@@ -1243,8 +1254,8 @@ static inline __u32 tcp_current_ssthresh(const struct sock *sk)
 		return tp->snd_ssthresh;
 	else
 		return max(tp->snd_ssthresh,
-			   ((tp->snd_cwnd >> 1) +
-			    (tp->snd_cwnd >> 2)));
+			   ((tcp_snd_cwnd(tp) >> 1) +
+			    (tcp_snd_cwnd(tp) >> 2)));
 }
 
 /* Use define here intentionally to get WARN_ON location shown at the caller */
@@ -1286,7 +1297,7 @@ static inline bool tcp_is_cwnd_limited(const struct sock *sk)
 
 	/* If in slow start, ensure cwnd grows to twice what was ACKed. */
 	if (tcp_in_slow_start(tp))
-		return tp->snd_cwnd < 2 * tp->max_packets_out;
+		return tcp_snd_cwnd(tp) < 2 * tp->max_packets_out;
 
 	return tp->is_cwnd_limited;
 }
include/trace/events/tcp.h +1 -1
index 521059d8dc0a..edcd6369de10 100644
@@ -279,7 +279,7 @@ TRACE_EVENT(tcp_probe,
 		__entry->data_len = skb->len - __tcp_hdrlen(th);
 		__entry->snd_nxt = tp->snd_nxt;
 		__entry->snd_una = tp->snd_una;
-		__entry->snd_cwnd = tp->snd_cwnd;
+		__entry->snd_cwnd = tcp_snd_cwnd(tp);
 		__entry->snd_wnd = tp->snd_wnd;
 		__entry->rcv_wnd = tp->rcv_wnd;
 		__entry->ssthresh = tcp_current_ssthresh(sk);
net/core/filter.c +1 -1
index f8fbb5fa74f3..4210b127c5f5 100644
@@ -4937,7 +4937,7 @@ static int _bpf_setsockopt(struct sock *sk, int level, int optname,
 				if (val <= 0 || tp->data_segs_out > tp->syn_data)
 					ret = -EINVAL;
 				else
-					tp->snd_cwnd = val;
+					tcp_snd_cwnd_set(tp, val);
 				break;
 			case TCP_BPF_SNDCWND_CLAMP:
 				if (val <= 0) {
net/ipv4/tcp.c +4 -4
index 28ff2a820f7c..c9ad372f8edb 100644
@@ -429,7 +429,7 @@ void tcp_init_sock(struct sock *sk)
 	 * algorithms that we must have the following bandaid to talk
 	 * efficiently to them.  -DaveM
 	 */
-	tp->snd_cwnd = TCP_INIT_CWND;
+	tcp_snd_cwnd_set(tp, TCP_INIT_CWND);
 
 	/* There's a bubble in the pipe until at least the first ACK. */
 	tp->app_limited = ~0U;
@@ -3033,7 +3033,7 @@ int tcp_disconnect(struct sock *sk, int flags)
 	icsk->icsk_rto_min = TCP_RTO_MIN;
 	icsk->icsk_delack_max = TCP_DELACK_MAX;
 	tp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
-	tp->snd_cwnd = TCP_INIT_CWND;
+	tcp_snd_cwnd_set(tp, TCP_INIT_CWND);
 	tp->snd_cwnd_cnt = 0;
 	tp->window_clamp = 0;
 	tp->delivered = 0;
@@ -3744,7 +3744,7 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info)
 	info->tcpi_max_pacing_rate = rate64;
 
 	info->tcpi_reordering = tp->reordering;
-	info->tcpi_snd_cwnd = tp->snd_cwnd;
+	info->tcpi_snd_cwnd = tcp_snd_cwnd(tp);
 
 	if (info->tcpi_state == TCP_LISTEN) {
 		/* listeners aliased fields :
@@ -3915,7 +3915,7 @@ struct sk_buff *tcp_get_timestamping_opt_stats(const struct sock *sk,
 	rate64 = tcp_compute_delivery_rate(tp);
 	nla_put_u64_64bit(stats, TCP_NLA_DELIVERY_RATE, rate64, TCP_NLA_PAD);
 
-	nla_put_u32(stats, TCP_NLA_SND_CWND, tp->snd_cwnd);
+	nla_put_u32(stats, TCP_NLA_SND_CWND, tcp_snd_cwnd(tp));
 	nla_put_u32(stats, TCP_NLA_REORDERING, tp->reordering);
 	nla_put_u32(stats, TCP_NLA_MIN_RTT, tcp_min_rtt(tp));
 
net/ipv4/tcp_bbr.c +10 -10
index ec5550089b4d..aefe12b7dbf7 100644
@@ -276,7 +276,7 @@ static void bbr_init_pacing_rate_from_rtt(struct sock *sk)
 	} else {			 /* no RTT sample yet */
 		rtt_us = USEC_PER_MSEC;	 /* use nominal default RTT */
 	}
-	bw = (u64)tp->snd_cwnd * BW_UNIT;
+	bw = (u64)tcp_snd_cwnd(tp) * BW_UNIT;
 	do_div(bw, rtt_us);
 	sk->sk_pacing_rate = bbr_bw_to_pacing_rate(sk, bw, bbr_high_gain);
 }
@@ -323,9 +323,9 @@ static void bbr_save_cwnd(struct sock *sk)
 	struct bbr *bbr = inet_csk_ca(sk);
 
 	if (bbr->prev_ca_state < TCP_CA_Recovery && bbr->mode != BBR_PROBE_RTT)
-		bbr->prior_cwnd = tp->snd_cwnd;  /* this cwnd is good enough */
+		bbr->prior_cwnd = tcp_snd_cwnd(tp);  /* this cwnd is good enough */
 	else  /* loss recovery or BBR_PROBE_RTT have temporarily cut cwnd */
-		bbr->prior_cwnd = max(bbr->prior_cwnd, tp->snd_cwnd);
+		bbr->prior_cwnd = max(bbr->prior_cwnd, tcp_snd_cwnd(tp));
 }
 
 static void bbr_cwnd_event(struct sock *sk, enum tcp_ca_event event)
@@ -482,7 +482,7 @@ static bool bbr_set_cwnd_to_recover_or_restore(
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bbr *bbr = inet_csk_ca(sk);
 	u8 prev_state = bbr->prev_ca_state, state = inet_csk(sk)->icsk_ca_state;
-	u32 cwnd = tp->snd_cwnd;
+	u32 cwnd = tcp_snd_cwnd(tp);
 
 	/* An ACK for P pkts should release at most 2*P packets. We do this
 	 * in two steps. First, here we deduct the number of lost packets.
@@ -520,7 +520,7 @@ static void bbr_set_cwnd(struct sock *sk, const struct rate_sample *rs,
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bbr *bbr = inet_csk_ca(sk);
-	u32 cwnd = tp->snd_cwnd, target_cwnd = 0;
+	u32 cwnd = tcp_snd_cwnd(tp), target_cwnd = 0;
 
 	if (!acked)
 		goto done;  /* no packet fully ACKed; just apply caps */
@@ -544,9 +544,9 @@ static void bbr_set_cwnd(struct sock *sk, const struct rate_sample *rs,
 	cwnd = max(cwnd, bbr_cwnd_min_target);
 
 done:
-	tp->snd_cwnd = min(cwnd, tp->snd_cwnd_clamp);	/* apply global cap */
+	tcp_snd_cwnd_set(tp, min(cwnd, tp->snd_cwnd_clamp));	/* apply global cap */
 	if (bbr->mode == BBR_PROBE_RTT)  /* drain queue, refresh min_rtt */
-		tp->snd_cwnd = min(tp->snd_cwnd, bbr_cwnd_min_target);
+		tcp_snd_cwnd_set(tp, min(tcp_snd_cwnd(tp), bbr_cwnd_min_target));
 }
 
 /* End cycle phase if it's time and/or we hit the phase's in-flight target. */
@@ -856,7 +856,7 @@ static void bbr_update_ack_aggregation(struct sock *sk,
 	bbr->ack_epoch_acked = min_t(u32, 0xFFFFF,
 				     bbr->ack_epoch_acked + rs->acked_sacked);
 	extra_acked = bbr->ack_epoch_acked - expected_acked;
-	extra_acked = min(extra_acked, tp->snd_cwnd);
+	extra_acked = min(extra_acked, tcp_snd_cwnd(tp));
 	if (extra_acked > bbr->extra_acked[bbr->extra_acked_win_idx])
 		bbr->extra_acked[bbr->extra_acked_win_idx] = extra_acked;
 }
@@ -914,7 +914,7 @@ static void bbr_check_probe_rtt_done(struct sock *sk)
 		return;
 
 	bbr->min_rtt_stamp = tcp_jiffies32;  /* wait a while until PROBE_RTT */
-	tp->snd_cwnd = max(tp->snd_cwnd, bbr->prior_cwnd);
+	tcp_snd_cwnd_set(tp, max(tcp_snd_cwnd(tp), bbr->prior_cwnd));
 	bbr_reset_mode(sk);
 }
 
@@ -1093,7 +1093,7 @@ static u32 bbr_undo_cwnd(struct sock *sk)
 	bbr->full_bw = 0;   /* spurious slow-down; reset full pipe detection */
 	bbr->full_bw_cnt = 0;
 	bbr_reset_lt_bw_sampling(sk);
-	return tcp_sk(sk)->snd_cwnd;
+	return tcp_snd_cwnd(tcp_sk(sk));
 }
 
 /* Entering loss recovery, so save cwnd for when we exit or undo recovery. */
net/ipv4/tcp_bic.c +7 -7
index f5f588b1f6e9..58358bf92e1b 100644
@@ -150,7 +150,7 @@ static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		if (!acked)
 			return;
 	}
-	bictcp_update(ca, tp->snd_cwnd);
+	bictcp_update(ca, tcp_snd_cwnd(tp));
 	tcp_cong_avoid_ai(tp, ca->cnt, acked);
 }
 
@@ -166,16 +166,16 @@ static u32 bictcp_recalc_ssthresh(struct sock *sk)
 	ca->epoch_start = 0;	/* end of epoch */
 
 	/* Wmax and fast convergence */
-	if (tp->snd_cwnd < ca->last_max_cwnd && fast_convergence)
-		ca->last_max_cwnd = (tp->snd_cwnd * (BICTCP_BETA_SCALE + beta))
+	if (tcp_snd_cwnd(tp) < ca->last_max_cwnd && fast_convergence)
+		ca->last_max_cwnd = (tcp_snd_cwnd(tp) * (BICTCP_BETA_SCALE + beta))
 			/ (2 * BICTCP_BETA_SCALE);
 	else
-		ca->last_max_cwnd = tp->snd_cwnd;
+		ca->last_max_cwnd = tcp_snd_cwnd(tp);
 
-	if (tp->snd_cwnd <= low_window)
-		return max(tp->snd_cwnd >> 1U, 2U);
+	if (tcp_snd_cwnd(tp) <= low_window)
+		return max(tcp_snd_cwnd(tp) >> 1U, 2U);
 	else
-		return max((tp->snd_cwnd * beta) / BICTCP_BETA_SCALE, 2U);
+		return max((tcp_snd_cwnd(tp) * beta) / BICTCP_BETA_SCALE, 2U);
 }
 
 static void bictcp_state(struct sock *sk, u8 new_state)
net/ipv4/tcp_cdg.c +15 -15
index 709d23801823..ddc7ba0554bd 100644
@@ -161,8 +161,8 @@ static void tcp_cdg_hystart_update(struct sock *sk)
 					      LINUX_MIB_TCPHYSTARTTRAINDETECT);
 				NET_ADD_STATS(sock_net(sk),
 					      LINUX_MIB_TCPHYSTARTTRAINCWND,
-					      tp->snd_cwnd);
-				tp->snd_ssthresh = tp->snd_cwnd;
+					      tcp_snd_cwnd(tp));
+				tp->snd_ssthresh = tcp_snd_cwnd(tp);
 				return;
 			}
 		}
@@ -180,8 +180,8 @@ static void tcp_cdg_hystart_update(struct sock *sk)
 					      LINUX_MIB_TCPHYSTARTDELAYDETECT);
 				NET_ADD_STATS(sock_net(sk),
 					      LINUX_MIB_TCPHYSTARTDELAYCWND,
-					      tp->snd_cwnd);
-				tp->snd_ssthresh = tp->snd_cwnd;
+					      tcp_snd_cwnd(tp));
+				tp->snd_ssthresh = tcp_snd_cwnd(tp);
 			}
 		}
 	}
@@ -252,7 +252,7 @@ static bool tcp_cdg_backoff(struct sock *sk, u32 grad)
 			return false;
 	}
 
-	ca->shadow_wnd = max(ca->shadow_wnd, tp->snd_cwnd);
+	ca->shadow_wnd = max(ca->shadow_wnd, tcp_snd_cwnd(tp));
 	ca->state = CDG_BACKOFF;
 	tcp_enter_cwr(sk);
 	return true;
@@ -285,14 +285,14 @@ static void tcp_cdg_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 	}
 
 	if (!tcp_is_cwnd_limited(sk)) {
-		ca->shadow_wnd = min(ca->shadow_wnd, tp->snd_cwnd);
+		ca->shadow_wnd = min(ca->shadow_wnd, tcp_snd_cwnd(tp));
 		return;
 	}
 
-	prior_snd_cwnd = tp->snd_cwnd;
+	prior_snd_cwnd = tcp_snd_cwnd(tp);
 	tcp_reno_cong_avoid(sk, ack, acked);
 
-	incr = tp->snd_cwnd - prior_snd_cwnd;
+	incr = tcp_snd_cwnd(tp) - prior_snd_cwnd;
 	ca->shadow_wnd = max(ca->shadow_wnd, ca->shadow_wnd + incr);
 }
 
@@ -331,15 +331,15 @@ static u32 tcp_cdg_ssthresh(struct sock *sk)
 	struct tcp_sock *tp = tcp_sk(sk);
 
 	if (ca->state == CDG_BACKOFF)
-		return max(2U, (tp->snd_cwnd * min(1024U, backoff_beta)) >> 10);
+		return max(2U, (tcp_snd_cwnd(tp) * min(1024U, backoff_beta)) >> 10);
 
 	if (ca->state == CDG_NONFULL && use_tolerance)
-		return tp->snd_cwnd;
+		return tcp_snd_cwnd(tp);
 
-	ca->shadow_wnd = min(ca->shadow_wnd >> 1, tp->snd_cwnd);
+	ca->shadow_wnd = min(ca->shadow_wnd >> 1, tcp_snd_cwnd(tp));
 	if (use_shadow)
-		return max3(2U, ca->shadow_wnd, tp->snd_cwnd >> 1);
-	return max(2U, tp->snd_cwnd >> 1);
+		return max3(2U, ca->shadow_wnd, tcp_snd_cwnd(tp) >> 1);
+	return max(2U, tcp_snd_cwnd(tp) >> 1);
 }
 
 static void tcp_cdg_cwnd_event(struct sock *sk, const enum tcp_ca_event ev)
@@ -357,7 +357,7 @@ static void tcp_cdg_cwnd_event(struct sock *sk, const enum tcp_ca_event ev)
 
 		ca->gradients = gradients;
 		ca->rtt_seq = tp->snd_nxt;
-		ca->shadow_wnd = tp->snd_cwnd;
+		ca->shadow_wnd = tcp_snd_cwnd(tp);
 		break;
 	case CA_EVENT_COMPLETE_CWR:
 		ca->state = CDG_UNKNOWN;
@@ -380,7 +380,7 @@ static void tcp_cdg_init(struct sock *sk)
 		ca->gradients = kcalloc(window, sizeof(ca->gradients[0]),
 					GFP_NOWAIT | __GFP_NOWARN);
 	ca->rtt_seq = tp->snd_nxt;
-	ca->shadow_wnd = tp->snd_cwnd;
+	ca->shadow_wnd = tcp_snd_cwnd(tp);
 }
 
 static void tcp_cdg_release(struct sock *sk)
net/ipv4/tcp_cong.c +9 -9
index db5831e6c136..f43db30a7195 100644
@@ -395,10 +395,10 @@ int tcp_set_congestion_control(struct sock *sk, const char *name, bool load,
  */
 u32 tcp_slow_start(struct tcp_sock *tp, u32 acked)
 {
-	u32 cwnd = min(tp->snd_cwnd + acked, tp->snd_ssthresh);
+	u32 cwnd = min(tcp_snd_cwnd(tp) + acked, tp->snd_ssthresh);
 
-	acked -= cwnd - tp->snd_cwnd;
-	tp->snd_cwnd = min(cwnd, tp->snd_cwnd_clamp);
+	acked -= cwnd - tcp_snd_cwnd(tp);
+	tcp_snd_cwnd_set(tp, min(cwnd, tp->snd_cwnd_clamp));
 
 	return acked;
 }
@@ -412,7 +412,7 @@ void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w, u32 acked)
 	/* If credits accumulated at a higher w, apply them gently now. */
 	if (tp->snd_cwnd_cnt >= w) {
 		tp->snd_cwnd_cnt = 0;
-		tp->snd_cwnd++;
+		tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 	}
 
 	tp->snd_cwnd_cnt += acked;
@@ -420,9 +420,9 @@ void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w, u32 acked)
 		u32 delta = tp->snd_cwnd_cnt / w;
 
 		tp->snd_cwnd_cnt -= delta * w;
-		tp->snd_cwnd += delta;
+		tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + delta);
 	}
-	tp->snd_cwnd = min(tp->snd_cwnd, tp->snd_cwnd_clamp);
+	tcp_snd_cwnd_set(tp, min(tcp_snd_cwnd(tp), tp->snd_cwnd_clamp));
 }
 EXPORT_SYMBOL_GPL(tcp_cong_avoid_ai);
 
@@ -447,7 +447,7 @@ void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 			return;
 	}
 	/* In dangerous area, increase slowly. */
-	tcp_cong_avoid_ai(tp, tp->snd_cwnd, acked);
+	tcp_cong_avoid_ai(tp, tcp_snd_cwnd(tp), acked);
 }
 EXPORT_SYMBOL_GPL(tcp_reno_cong_avoid);
 
@@ -456,7 +456,7 @@ u32 tcp_reno_ssthresh(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 
-	return max(tp->snd_cwnd >> 1U, 2U);
+	return max(tcp_snd_cwnd(tp) >> 1U, 2U);
 }
 EXPORT_SYMBOL_GPL(tcp_reno_ssthresh);
 
@@ -464,7 +464,7 @@ u32 tcp_reno_undo_cwnd(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 
-	return max(tp->snd_cwnd, tp->prior_cwnd);
+	return max(tcp_snd_cwnd(tp), tp->prior_cwnd);
 }
 EXPORT_SYMBOL_GPL(tcp_reno_undo_cwnd);
 
net/ipv4/tcp_cubic.c +11 -11
index e07837e23b3f..f0a240792ff9 100644
@@ -334,7 +334,7 @@ static void cubictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		if (!acked)
 			return;
 	}
-	bictcp_update(ca, tp->snd_cwnd, acked);
+	bictcp_update(ca, tcp_snd_cwnd(tp), acked);
 	tcp_cong_avoid_ai(tp, ca->cnt, acked);
 }
 
@@ -346,13 +346,13 @@ static u32 cubictcp_recalc_ssthresh(struct sock *sk)
 	ca->epoch_start = 0;	/* end of epoch */
 
 	/* Wmax and fast convergence */
-	if (tp->snd_cwnd < ca->last_max_cwnd && fast_convergence)
-		ca->last_max_cwnd = (tp->snd_cwnd * (BICTCP_BETA_SCALE + beta))
+	if (tcp_snd_cwnd(tp) < ca->last_max_cwnd && fast_convergence)
+		ca->last_max_cwnd = (tcp_snd_cwnd(tp) * (BICTCP_BETA_SCALE + beta))
 			/ (2 * BICTCP_BETA_SCALE);
 	else
-		ca->last_max_cwnd = tp->snd_cwnd;
+		ca->last_max_cwnd = tcp_snd_cwnd(tp);
 
-	return max((tp->snd_cwnd * beta) / BICTCP_BETA_SCALE, 2U);
+	return max((tcp_snd_cwnd(tp) * beta) / BICTCP_BETA_SCALE, 2U);
 }
 
 static void cubictcp_state(struct sock *sk, u8 new_state)
@@ -413,13 +413,13 @@ static void hystart_update(struct sock *sk, u32 delay)
 				ca->found = 1;
 				pr_debug("hystart_ack_train (%u > %u) delay_min %u (+ ack_delay %u) cwnd %u\n",
 					 now - ca->round_start, threshold,
-					 ca->delay_min, hystart_ack_delay(sk), tp->snd_cwnd);
+					 ca->delay_min, hystart_ack_delay(sk), tcp_snd_cwnd(tp));
 				NET_INC_STATS(sock_net(sk),
 					      LINUX_MIB_TCPHYSTARTTRAINDETECT);
 				NET_ADD_STATS(sock_net(sk),
 					      LINUX_MIB_TCPHYSTARTTRAINCWND,
-					      tp->snd_cwnd);
-				tp->snd_ssthresh = tp->snd_cwnd;
+					      tcp_snd_cwnd(tp));
+				tp->snd_ssthresh = tcp_snd_cwnd(tp);
 			}
 		}
 	}
@@ -438,8 +438,8 @@ static void hystart_update(struct sock *sk, u32 delay)
 					      LINUX_MIB_TCPHYSTARTDELAYDETECT);
 				NET_ADD_STATS(sock_net(sk),
 					      LINUX_MIB_TCPHYSTARTDELAYCWND,
-					      tp->snd_cwnd);
-				tp->snd_ssthresh = tp->snd_cwnd;
+					      tcp_snd_cwnd(tp));
+				tp->snd_ssthresh = tcp_snd_cwnd(tp);
 			}
 		}
 	}
@@ -469,7 +469,7 @@ static void cubictcp_acked(struct sock *sk, const struct ack_sample *sample)
 
 	/* hystart triggers when cwnd is larger than some threshold */
 	if (!ca->found && tcp_in_slow_start(tp) && hystart &&
-	    tp->snd_cwnd >= hystart_low_window)
+	    tcp_snd_cwnd(tp) >= hystart_low_window)
 		hystart_update(sk, delay);
 }
 
net/ipv4/tcp_dctcp.c +6 -5
index 0d7ab3cc7b61..d0bf7bb4b140 100644
@@ -106,8 +106,8 @@ static u32 dctcp_ssthresh(struct sock *sk)
 	struct dctcp *ca = inet_csk_ca(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	ca->loss_cwnd = tp->snd_cwnd;
-	return max(tp->snd_cwnd - ((tp->snd_cwnd * ca->dctcp_alpha) >> 11U), 2U);
+	ca->loss_cwnd = tcp_snd_cwnd(tp);
+	return max(tcp_snd_cwnd(tp) - ((tcp_snd_cwnd(tp) * ca->dctcp_alpha) >> 11U), 2U);
 }
 
 static void dctcp_update_alpha(struct sock *sk, u32 flags)
@@ -148,8 +148,8 @@ static void dctcp_react_to_loss(struct sock *sk)
 	struct dctcp *ca = inet_csk_ca(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	ca->loss_cwnd = tp->snd_cwnd;
-	tp->snd_ssthresh = max(tp->snd_cwnd >> 1U, 2U);
+	ca->loss_cwnd = tcp_snd_cwnd(tp);
+	tp->snd_ssthresh = max(tcp_snd_cwnd(tp) >> 1U, 2U);
 }
 
 static void dctcp_state(struct sock *sk, u8 new_state)
@@ -211,8 +211,9 @@ static size_t dctcp_get_info(struct sock *sk, u32 ext, int *attr,
 static u32 dctcp_cwnd_undo(struct sock *sk)
 {
 	const struct dctcp *ca = inet_csk_ca(sk);
+	struct tcp_sock *tp = tcp_sk(sk);
 
-	return max(tcp_sk(sk)->snd_cwnd, ca->loss_cwnd);
+	return max(tcp_snd_cwnd(tp), ca->loss_cwnd);
 }
 
 static struct tcp_congestion_ops dctcp __read_mostly = {
net/ipv4/tcp_highspeed.c +9 -9
index 349069d6cd0a..c6de5ce79ad3 100644
@@ -127,22 +127,22 @@ static void hstcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		 *     snd_cwnd <=
 		 *     hstcp_aimd_vals[ca->ai].cwnd
 		 */
-		if (tp->snd_cwnd > hstcp_aimd_vals[ca->ai].cwnd) {
-			while (tp->snd_cwnd > hstcp_aimd_vals[ca->ai].cwnd &&
+		if (tcp_snd_cwnd(tp) > hstcp_aimd_vals[ca->ai].cwnd) {
+			while (tcp_snd_cwnd(tp) > hstcp_aimd_vals[ca->ai].cwnd &&
 			       ca->ai < HSTCP_AIMD_MAX - 1)
 				ca->ai++;
-		} else if (ca->ai && tp->snd_cwnd <= hstcp_aimd_vals[ca->ai-1].cwnd) {
-			while (ca->ai && tp->snd_cwnd <= hstcp_aimd_vals[ca->ai-1].cwnd)
+		} else if (ca->ai && tcp_snd_cwnd(tp) <= hstcp_aimd_vals[ca->ai-1].cwnd) {
+			while (ca->ai && tcp_snd_cwnd(tp) <= hstcp_aimd_vals[ca->ai-1].cwnd)
 				ca->ai--;
 		}
 
 		/* Do additive increase */
-		if (tp->snd_cwnd < tp->snd_cwnd_clamp) {
+		if (tcp_snd_cwnd(tp) < tp->snd_cwnd_clamp) {
 			/* cwnd = cwnd + a(w) / cwnd */
 			tp->snd_cwnd_cnt += ca->ai + 1;
-			if (tp->snd_cwnd_cnt >= tp->snd_cwnd) {
-				tp->snd_cwnd_cnt -= tp->snd_cwnd;
-				tp->snd_cwnd++;
+			if (tp->snd_cwnd_cnt >= tcp_snd_cwnd(tp)) {
+				tp->snd_cwnd_cnt -= tcp_snd_cwnd(tp);
+				tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 			}
 		}
 	}
@@ -154,7 +154,7 @@ static u32 hstcp_ssthresh(struct sock *sk)
 	struct hstcp *ca = inet_csk_ca(sk);
 
 	/* Do multiplicative decrease */
-	return max(tp->snd_cwnd - ((tp->snd_cwnd * hstcp_aimd_vals[ca->ai].md) >> 8), 2U);
+	return max(tcp_snd_cwnd(tp) - ((tcp_snd_cwnd(tp) * hstcp_aimd_vals[ca->ai].md) >> 8), 2U);
 }
 
 static struct tcp_congestion_ops tcp_highspeed __read_mostly = {
net/ipv4/tcp_htcp.c +5 -5
index 55adcfcf96fe..52b1f2665dfa 100644
@@ -124,7 +124,7 @@ static void measure_achieved_throughput(struct sock *sk,
 
 	ca->packetcount += sample->pkts_acked;
 
-	if (ca->packetcount >= tp->snd_cwnd - (ca->alpha >> 7 ? : 1) &&
+	if (ca->packetcount >= tcp_snd_cwnd(tp) - (ca->alpha >> 7 ? : 1) &&
 	    now - ca->lasttime >= ca->minRTT &&
 	    ca->minRTT > 0) {
 		__u32 cur_Bi = ca->packetcount * HZ / (now - ca->lasttime);
@@ -225,7 +225,7 @@ static u32 htcp_recalc_ssthresh(struct sock *sk)
 	const struct htcp *ca = inet_csk_ca(sk);
 
 	htcp_param_update(sk);
-	return max((tp->snd_cwnd * ca->beta) >> 7, 2U);
+	return max((tcp_snd_cwnd(tp) * ca->beta) >> 7, 2U);
 }
 
 static void htcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
@@ -242,9 +242,9 @@ static void htcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		/* In dangerous area, increase slowly.
 		 * In theory this is tp->snd_cwnd += alpha / tp->snd_cwnd
 		 */
-		if ((tp->snd_cwnd_cnt * ca->alpha)>>7 >= tp->snd_cwnd) {
-			if (tp->snd_cwnd < tp->snd_cwnd_clamp)
-				tp->snd_cwnd++;
+		if ((tp->snd_cwnd_cnt * ca->alpha)>>7 >= tcp_snd_cwnd(tp)) {
+			if (tcp_snd_cwnd(tp) < tp->snd_cwnd_clamp)
+				tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 			tp->snd_cwnd_cnt = 0;
 			htcp_alpha_update(ca);
 		} else
net/ipv4/tcp_hybla.c +9 -9
index be39327e04e6..abd7d91807e5 100644
@@ -54,7 +54,7 @@ static void hybla_init(struct sock *sk)
 	ca->rho2_7ls = 0;
 	ca->snd_cwnd_cents = 0;
 	ca->hybla_en = true;
-	tp->snd_cwnd = 2;
+	tcp_snd_cwnd_set(tp, 2);
 	tp->snd_cwnd_clamp = 65535;
 
 	/* 1st Rho measurement based on initial srtt */
@@ -62,7 +62,7 @@ static void hybla_init(struct sock *sk)
 
 	/* set minimum rtt as this is the 1st ever seen */
 	ca->minrtt_us = tp->srtt_us;
-	tp->snd_cwnd = ca->rho;
+	tcp_snd_cwnd_set(tp, ca->rho);
 }
 
 static void hybla_state(struct sock *sk, u8 ca_state)
@@ -137,31 +137,31 @@ static void hybla_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		 * as long as increment is estimated as (rho<<7)/window
 		 * it already is <<7 and we can easily count its fractions.
 		 */
-		increment = ca->rho2_7ls / tp->snd_cwnd;
+		increment = ca->rho2_7ls / tcp_snd_cwnd(tp);
 		if (increment < 128)
 			tp->snd_cwnd_cnt++;
 	}
 
 	odd = increment % 128;
-	tp->snd_cwnd += increment >> 7;
+	tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + (increment >> 7));
 	ca->snd_cwnd_cents += odd;
 
 	/* check when fractions goes >=128 and increase cwnd by 1. */
 	while (ca->snd_cwnd_cents >= 128) {
-		tp->snd_cwnd++;
+		tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 		ca->snd_cwnd_cents -= 128;
 		tp->snd_cwnd_cnt = 0;
 	}
 	/* check when cwnd has not been incremented for a while */
-	if (increment == 0 && odd == 0 && tp->snd_cwnd_cnt >= tp->snd_cwnd) {
-		tp->snd_cwnd++;
+	if (increment == 0 && odd == 0 && tp->snd_cwnd_cnt >= tcp_snd_cwnd(tp)) {
+		tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 		tp->snd_cwnd_cnt = 0;
 	}
 	/* clamp down slowstart cwnd to ssthresh value. */
 	if (is_slowstart)
-		tp->snd_cwnd = min(tp->snd_cwnd, tp->snd_ssthresh);
+		tcp_snd_cwnd_set(tp, min(tcp_snd_cwnd(tp), tp->snd_ssthresh));
 
-	tp->snd_cwnd = min_t(u32, tp->snd_cwnd, tp->snd_cwnd_clamp);
+	tcp_snd_cwnd_set(tp, min(tcp_snd_cwnd(tp), tp->snd_cwnd_clamp));
 }
 
 static struct tcp_congestion_ops tcp_hybla __read_mostly = {
net/ipv4/tcp_illinois.c +7 -5
index 00e54873213e..c0c81a2c77fa 100644
@@ -224,7 +224,7 @@ static void update_params(struct sock *sk)
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct illinois *ca = inet_csk_ca(sk);
 
-	if (tp->snd_cwnd < win_thresh) {
+	if (tcp_snd_cwnd(tp) < win_thresh) {
 		ca->alpha = ALPHA_BASE;
 		ca->beta = BETA_BASE;
 	} else if (ca->cnt_rtt > 0) {
@@ -284,9 +284,9 @@ static void tcp_illinois_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		 * tp->snd_cwnd += alpha/tp->snd_cwnd
 		*/
 		delta = (tp->snd_cwnd_cnt * ca->alpha) >> ALPHA_SHIFT;
-		if (delta >= tp->snd_cwnd) {
-			tp->snd_cwnd = min(tp->snd_cwnd + delta / tp->snd_cwnd,
-					   (u32)tp->snd_cwnd_clamp);
+		if (delta >= tcp_snd_cwnd(tp)) {
+			tcp_snd_cwnd_set(tp, min(tcp_snd_cwnd(tp) + delta / tcp_snd_cwnd(tp),
+						 (u32)tp->snd_cwnd_clamp));
 			tp->snd_cwnd_cnt = 0;
 		}
 	}
@@ -296,9 +296,11 @@ static u32 tcp_illinois_ssthresh(struct sock *sk)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct illinois *ca = inet_csk_ca(sk);
+	u32 decr;
 
 	/* Multiplicative decrease */
-	return max(tp->snd_cwnd - ((tp->snd_cwnd * ca->beta) >> BETA_SHIFT), 2U);
+	decr = (tcp_snd_cwnd(tp) * ca->beta) >> BETA_SHIFT;
+	return max(tcp_snd_cwnd(tp) - decr, 2U);
 }
 
 /* Extract info for Tcp socket info provided via netlink. */
net/ipv4/tcp_input.c +18 -18
index 96c25c97ee56..edd952250465 100644
@@ -414,7 +414,7 @@ static void tcp_sndbuf_expand(struct sock *sk)
 	per_mss = roundup_pow_of_two(per_mss) +
 		  SKB_DATA_ALIGN(sizeof(struct sk_buff));
 
-	nr_segs = max_t(u32, TCP_INIT_CWND, tp->snd_cwnd);
+	nr_segs = max_t(u32, TCP_INIT_CWND, tcp_snd_cwnd(tp));
 	nr_segs = max_t(u32, nr_segs, tp->reordering + 1);
 
 	/* Fast Recovery (RFC 5681 3.2) :
@@ -909,12 +909,12 @@ static void tcp_update_pacing_rate(struct sock *sk)
 	 *	 If snd_cwnd >= (tp->snd_ssthresh / 2), we are approaching
 	 *	 end of slow start and should slow down.
 	 */
-	if (tp->snd_cwnd < tp->snd_ssthresh / 2)
+	if (tcp_snd_cwnd(tp) < tp->snd_ssthresh / 2)
 		rate *= sock_net(sk)->ipv4.sysctl_tcp_pacing_ss_ratio;
 	else
 		rate *= sock_net(sk)->ipv4.sysctl_tcp_pacing_ca_ratio;
 
-	rate *= max(tp->snd_cwnd, tp->packets_out);
+	rate *= max(tcp_snd_cwnd(tp), tp->packets_out);
 
 	if (likely(tp->srtt_us))
 		do_div(rate, tp->srtt_us);
@@ -2147,12 +2147,12 @@ void tcp_enter_loss(struct sock *sk)
 	    !after(tp->high_seq, tp->snd_una) ||
 	    (icsk->icsk_ca_state == TCP_CA_Loss && !icsk->icsk_retransmits)) {
 		tp->prior_ssthresh = tcp_current_ssthresh(sk);
-		tp->prior_cwnd = tp->snd_cwnd;
+		tp->prior_cwnd = tcp_snd_cwnd(tp);
 		tp->snd_ssthresh = icsk->icsk_ca_ops->ssthresh(sk);
 		tcp_ca_event(sk, CA_EVENT_LOSS);
 		tcp_init_undo(tp);
 	}
-	tp->snd_cwnd	   = tcp_packets_in_flight(tp) + 1;
+	tcp_snd_cwnd_set(tp, tcp_packets_in_flight(tp) + 1);
 	tp->snd_cwnd_cnt   = 0;
 	tp->snd_cwnd_stamp = tcp_jiffies32;
 
@@ -2458,7 +2458,7 @@ static void DBGUNDO(struct sock *sk, const char *msg)
 		pr_debug("Undo %s %pI4/%u c%u l%u ss%u/%u p%u\n",
 			 msg,
 			 &inet->inet_daddr, ntohs(inet->inet_dport),
-			 tp->snd_cwnd, tcp_left_out(tp),
+			 tcp_snd_cwnd(tp), tcp_left_out(tp),
 			 tp->snd_ssthresh, tp->prior_ssthresh,
 			 tp->packets_out);
 	}
@@ -2467,7 +2467,7 @@ static void DBGUNDO(struct sock *sk, const char *msg)
 		pr_debug("Undo %s %pI6/%u c%u l%u ss%u/%u p%u\n",
 			 msg,
 			 &sk->sk_v6_daddr, ntohs(inet->inet_dport),
-			 tp->snd_cwnd, tcp_left_out(tp),
+			 tcp_snd_cwnd(tp), tcp_left_out(tp),
 			 tp->snd_ssthresh, tp->prior_ssthresh,
 			 tp->packets_out);
 	}
@@ -2492,7 +2492,7 @@ static void tcp_undo_cwnd_reduction(struct sock *sk, bool unmark_loss)
 	if (tp->prior_ssthresh) {
 		const struct inet_connection_sock *icsk = inet_csk(sk);
 
-		tp->snd_cwnd = icsk->icsk_ca_ops->undo_cwnd(sk);
+		tcp_snd_cwnd_set(tp, icsk->icsk_ca_ops->undo_cwnd(sk));
 
 		if (tp->prior_ssthresh > tp->snd_ssthresh) {
 			tp->snd_ssthresh = tp->prior_ssthresh;
@@ -2599,7 +2599,7 @@ static void tcp_init_cwnd_reduction(struct sock *sk)
 	tp->high_seq = tp->snd_nxt;
 	tp->tlp_high_seq = 0;
 	tp->snd_cwnd_cnt = 0;
-	tp->prior_cwnd = tp->snd_cwnd;
+	tp->prior_cwnd = tcp_snd_cwnd(tp);
 	tp->prr_delivered = 0;
 	tp->prr_out = 0;
 	tp->snd_ssthresh = inet_csk(sk)->icsk_ca_ops->ssthresh(sk);
@@ -2629,7 +2629,7 @@ void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int newly_lost,
 	}
 	/* Force a fast retransmit upon entering fast recovery */
 	sndcnt = max(sndcnt, (tp->prr_out ? 0 : 1));
-	tp->snd_cwnd = tcp_packets_in_flight(tp) + sndcnt;
+	tcp_snd_cwnd_set(tp, tcp_packets_in_flight(tp) + sndcnt);
 }
 
 static inline void tcp_end_cwnd_reduction(struct sock *sk)
@@ -2642,7 +2642,7 @@ static inline void tcp_end_cwnd_reduction(struct sock *sk)
 	/* Reset cwnd to ssthresh in CWR or Recovery (unless it's undone) */
 	if (tp->snd_ssthresh < TCP_INFINITE_SSTHRESH &&
 	    (inet_csk(sk)->icsk_ca_state == TCP_CA_CWR || tp->undo_marker)) {
-		tp->snd_cwnd = tp->snd_ssthresh;
+		tcp_snd_cwnd_set(tp, tp->snd_ssthresh);
 		tp->snd_cwnd_stamp = tcp_jiffies32;
 	}
 	tcp_ca_event(sk, CA_EVENT_COMPLETE_CWR);
@@ -2709,9 +2709,9 @@ static void tcp_mtup_probe_success(struct sock *sk)
 
 	/* FIXME: breaks with very large cwnd */
 	tp->prior_ssthresh = tcp_current_ssthresh(sk);
-	tp->snd_cwnd = tp->snd_cwnd *
-		       tcp_mss_to_mtu(sk, tp->mss_cache) /
-		       icsk->icsk_mtup.probe_size;
+	tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) *
+			     tcp_mss_to_mtu(sk, tp->mss_cache) /
+			     icsk->icsk_mtup.probe_size);
 	tp->snd_cwnd_cnt = 0;
 	tp->snd_cwnd_stamp = tcp_jiffies32;
 	tp->snd_ssthresh = tcp_current_ssthresh(sk);
@@ -3034,7 +3034,7 @@ static void tcp_fastretrans_alert(struct sock *sk, const u32 prior_snd_una,
 		    tp->snd_una == tp->mtu_probe.probe_seq_start) {
 			tcp_mtup_probe_failed(sk);
 			/* Restores the reduction we did in tcp_mtup_probe() */
-			tp->snd_cwnd++;
+			tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 			tcp_simple_retransmit(sk);
 			return;
 		}
@@ -5420,7 +5420,7 @@ static bool tcp_should_expand_sndbuf(struct sock *sk)
 		return false;
 
 	/* If we filled the congestion window, do not expand.  */
-	if (tcp_packets_in_flight(tp) >= tp->snd_cwnd)
+	if (tcp_packets_in_flight(tp) >= tcp_snd_cwnd(tp))
 		return false;
 
 	return true;
@@ -5991,9 +5991,9 @@ void tcp_init_transfer(struct sock *sk, int bpf_op, struct sk_buff *skb)
 	 * retransmission has occurred.
 	 */
 	if (tp->total_retrans > 1 && tp->undo_marker)
-		tp->snd_cwnd = 1;
+		tcp_snd_cwnd_set(tp, 1);
 	else
-		tp->snd_cwnd = tcp_init_cwnd(tp, __sk_dst_get(sk));
+		tcp_snd_cwnd_set(tp, tcp_init_cwnd(tp, __sk_dst_get(sk)));
 	tp->snd_cwnd_stamp = tcp_jiffies32;
 
 	bpf_skops_established(sk, bpf_op, skb);
net/ipv4/tcp_ipv4.c +1 -1
index fec656f5a39e..79f9a6187a01 100644
@@ -2665,7 +2665,7 @@ static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i)
 		jiffies_to_clock_t(icsk->icsk_rto),
 		jiffies_to_clock_t(icsk->icsk_ack.ato),
 		(icsk->icsk_ack.quick << 1) | inet_csk_in_pingpong_mode(sk),
-		tp->snd_cwnd,
+		tcp_snd_cwnd(tp),
 		state == TCP_LISTEN ?
 		    fastopenq->max_qlen :
 		    (tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh));
net/ipv4/tcp_lp.c +3 -3
index 82b36ec3f2f8..ae36780977d2 100644
@@ -297,7 +297,7 @@ static void tcp_lp_pkts_acked(struct sock *sk, const struct ack_sample *sample)
 		lp->flag &= ~LP_WITHIN_THR;
 
 	pr_debug("TCP-LP: %05o|%5u|%5u|%15u|%15u|%15u\n", lp->flag,
-		 tp->snd_cwnd, lp->remote_hz, lp->owd_min, lp->owd_max,
+		 tcp_snd_cwnd(tp), lp->remote_hz, lp->owd_min, lp->owd_max,
 		 lp->sowd >> 3);
 
 	if (lp->flag & LP_WITHIN_THR)
@@ -313,12 +313,12 @@ static void tcp_lp_pkts_acked(struct sock *sk, const struct ack_sample *sample)
 	/* happened within inference
 	 * drop snd_cwnd into 1 */
 	if (lp->flag & LP_WITHIN_INF)
-		tp->snd_cwnd = 1U;
+		tcp_snd_cwnd_set(tp, 1U);
 
 	/* happened after inference
 	 * cut snd_cwnd into half */
 	else
-		tp->snd_cwnd = max(tp->snd_cwnd >> 1U, 1U);
+		tcp_snd_cwnd_set(tp, max(tcp_snd_cwnd(tp) >> 1U, 1U));
 
 	/* record this drop time */
 	lp->last_drop = now;
net/ipv4/tcp_metrics.c +6 -6
index 0588b004ddac..7029b0e98edb 100644
@@ -388,15 +388,15 @@ void tcp_update_metrics(struct sock *sk)
 		if (!net->ipv4.sysctl_tcp_no_ssthresh_metrics_save &&
 		    !tcp_metric_locked(tm, TCP_METRIC_SSTHRESH)) {
 			val = tcp_metric_get(tm, TCP_METRIC_SSTHRESH);
-			if (val && (tp->snd_cwnd >> 1) > val)
+			if (val && (tcp_snd_cwnd(tp) >> 1) > val)
 				tcp_metric_set(tm, TCP_METRIC_SSTHRESH,
-					       tp->snd_cwnd >> 1);
+					       tcp_snd_cwnd(tp) >> 1);
 		}
 		if (!tcp_metric_locked(tm, TCP_METRIC_CWND)) {
 			val = tcp_metric_get(tm, TCP_METRIC_CWND);
-			if (tp->snd_cwnd > val)
+			if (tcp_snd_cwnd(tp) > val)
 				tcp_metric_set(tm, TCP_METRIC_CWND,
-					       tp->snd_cwnd);
+					       tcp_snd_cwnd(tp));
 		}
 	} else if (!tcp_in_slow_start(tp) &&
 		   icsk->icsk_ca_state == TCP_CA_Open) {
@@ -404,10 +404,10 @@ void tcp_update_metrics(struct sock *sk)
 		if (!net->ipv4.sysctl_tcp_no_ssthresh_metrics_save &&
 		    !tcp_metric_locked(tm, TCP_METRIC_SSTHRESH))
 			tcp_metric_set(tm, TCP_METRIC_SSTHRESH,
-				       max(tp->snd_cwnd >> 1, tp->snd_ssthresh));
+				       max(tcp_snd_cwnd(tp) >> 1, tp->snd_ssthresh));
 		if (!tcp_metric_locked(tm, TCP_METRIC_CWND)) {
 			val = tcp_metric_get(tm, TCP_METRIC_CWND);
-			tcp_metric_set(tm, TCP_METRIC_CWND, (val + tp->snd_cwnd) >> 1);
+			tcp_metric_set(tm, TCP_METRIC_CWND, (val + tcp_snd_cwnd(tp)) >> 1);
 		}
 	} else {
 		/* Else slow start did not finish, cwnd is non-sense,
net/ipv4/tcp_nv.c +12 -12
index ab552356bdba..a60662f4bdf9 100644
@@ -197,10 +197,10 @@ static void tcpnv_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 	}
 
 	if (ca->cwnd_growth_factor < 0) {
-		cnt = tp->snd_cwnd << -ca->cwnd_growth_factor;
+		cnt = tcp_snd_cwnd(tp) << -ca->cwnd_growth_factor;
 		tcp_cong_avoid_ai(tp, cnt, acked);
 	} else {
-		cnt = max(4U, tp->snd_cwnd >> ca->cwnd_growth_factor);
+		cnt = max(4U, tcp_snd_cwnd(tp) >> ca->cwnd_growth_factor);
 		tcp_cong_avoid_ai(tp, cnt, acked);
 	}
 }
@@ -209,7 +209,7 @@ static u32 tcpnv_recalc_ssthresh(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 
-	return max((tp->snd_cwnd * nv_loss_dec_factor) >> 10, 2U);
+	return max((tcp_snd_cwnd(tp) * nv_loss_dec_factor) >> 10, 2U);
 }
 
 static void tcpnv_state(struct sock *sk, u8 new_state)
@@ -257,7 +257,7 @@ static void tcpnv_acked(struct sock *sk, const struct ack_sample *sample)
 		return;
 
 	/* Stop cwnd growth if we were in catch up mode */
-	if (ca->nv_catchup && tp->snd_cwnd >= nv_min_cwnd) {
+	if (ca->nv_catchup && tcp_snd_cwnd(tp) >= nv_min_cwnd) {
 		ca->nv_catchup = 0;
 		ca->nv_allow_cwnd_growth = 0;
 	}
@@ -371,7 +371,7 @@ static void tcpnv_acked(struct sock *sk, const struct ack_sample *sample)
 		 * if cwnd < max_win, grow cwnd
 		 * else leave the same
 		 */
-		if (tp->snd_cwnd > max_win) {
+		if (tcp_snd_cwnd(tp) > max_win) {
 			/* there is congestion, check that it is ok
 			 * to make a CA decision
 			 * 1. We should have at least nv_dec_eval_min_calls
@@ -398,20 +398,20 @@ static void tcpnv_acked(struct sock *sk, const struct ack_sample *sample)
 			ca->nv_allow_cwnd_growth = 0;
 			tp->snd_ssthresh =
 				(nv_ssthresh_factor * max_win) >> 3;
-			if (tp->snd_cwnd - max_win > 2) {
+			if (tcp_snd_cwnd(tp) - max_win > 2) {
 				/* gap > 2, we do exponential cwnd decrease */
 				int dec;
 
-				dec = max(2U, ((tp->snd_cwnd - max_win) *
+				dec = max(2U, ((tcp_snd_cwnd(tp) - max_win) *
 					       nv_cong_dec_mult) >> 7);
-				tp->snd_cwnd -= dec;
+				tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) - dec);
 			} else if (nv_cong_dec_mult > 0) {
-				tp->snd_cwnd = max_win;
+				tcp_snd_cwnd_set(tp, max_win);
 			}
 			if (ca->cwnd_growth_factor > 0)
 				ca->cwnd_growth_factor = 0;
 			ca->nv_no_cong_cnt = 0;
-		} else if (tp->snd_cwnd <= max_win - nv_pad_buffer) {
+		} else if (tcp_snd_cwnd(tp) <= max_win - nv_pad_buffer) {
 			/* There is no congestion, grow cwnd if allowed*/
 			if (ca->nv_eval_call_cnt < nv_inc_eval_min_calls)
 				return;
@@ -444,8 +444,8 @@ static void tcpnv_acked(struct sock *sk, const struct ack_sample *sample)
 		 * (it wasn't before, if it is now is because nv
 		 *  decreased it).
 		 */
-		if (tp->snd_cwnd < nv_min_cwnd)
-			tp->snd_cwnd = nv_min_cwnd;
+		if (tcp_snd_cwnd(tp) < nv_min_cwnd)
+			tcp_snd_cwnd_set(tp, nv_min_cwnd);
 	}
 }
 
net/ipv4/tcp_output.c +15 -15
index 0b5eab685154..84651fc3dde0 100644
@@ -143,7 +143,7 @@ void tcp_cwnd_restart(struct sock *sk, s32 delta)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	u32 restart_cwnd = tcp_init_cwnd(tp, __sk_dst_get(sk));
-	u32 cwnd = tp->snd_cwnd;
+	u32 cwnd = tcp_snd_cwnd(tp);
 
 	tcp_ca_event(sk, CA_EVENT_CWND_RESTART);
 
@@ -152,7 +152,7 @@ void tcp_cwnd_restart(struct sock *sk, s32 delta)
 
 	while ((delta -= inet_csk(sk)->icsk_rto) > 0 && cwnd > restart_cwnd)
 		cwnd >>= 1;
-	tp->snd_cwnd = max(cwnd, restart_cwnd);
+	tcp_snd_cwnd_set(tp, max(cwnd, restart_cwnd));
 	tp->snd_cwnd_stamp = tcp_jiffies32;
 	tp->snd_cwnd_used = 0;
 }
@@ -1014,7 +1014,7 @@ static void tcp_tsq_write(struct sock *sk)
 		struct tcp_sock *tp = tcp_sk(sk);
 
 		if (tp->lost_out > tp->retrans_out &&
-		    tp->snd_cwnd > tcp_packets_in_flight(tp)) {
+		    tcp_snd_cwnd(tp) > tcp_packets_in_flight(tp)) {
 			tcp_mstamp_refresh(tp);
 			tcp_xmit_retransmit_queue(sk);
 		}
@@ -1861,9 +1861,9 @@ static void tcp_cwnd_application_limited(struct sock *sk)
 		/* Limited by application or receiver window. */
 		u32 init_win = tcp_init_cwnd(tp, __sk_dst_get(sk));
 		u32 win_used = max(tp->snd_cwnd_used, init_win);
-		if (win_used < tp->snd_cwnd) {
+		if (win_used < tcp_snd_cwnd(tp)) {
 			tp->snd_ssthresh = tcp_current_ssthresh(sk);
-			tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1;
+			tcp_snd_cwnd_set(tp, (tcp_snd_cwnd(tp) + win_used) >> 1);
 		}
 		tp->snd_cwnd_used = 0;
 	}
@@ -2035,7 +2035,7 @@ static inline unsigned int tcp_cwnd_test(const struct tcp_sock *tp,
 		return 1;
 
 	in_flight = tcp_packets_in_flight(tp);
-	cwnd = tp->snd_cwnd;
+	cwnd = tcp_snd_cwnd(tp);
 	if (in_flight >= cwnd)
 		return 0;
 
@@ -2188,12 +2188,12 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb,
 	in_flight = tcp_packets_in_flight(tp);
 
 	BUG_ON(tcp_skb_pcount(skb) <= 1);
-	BUG_ON(tp->snd_cwnd <= in_flight);
+	BUG_ON(tcp_snd_cwnd(tp) <= in_flight);
 
 	send_win = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;
 
 	/* From in_flight test above, we know that cwnd > in_flight.  */
-	cong_win = (tp->snd_cwnd - in_flight) * tp->mss_cache;
+	cong_win = (tcp_snd_cwnd(tp) - in_flight) * tp->mss_cache;
 
 	limit = min(send_win, cong_win);
 
@@ -2207,7 +2207,7 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb,
 
 	win_divisor = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_tso_win_divisor);
 	if (win_divisor) {
-		u32 chunk = min(tp->snd_wnd, tp->snd_cwnd * tp->mss_cache);
+		u32 chunk = min(tp->snd_wnd, tcp_snd_cwnd(tp) * tp->mss_cache);
 
 		/* If at least some fraction of a window is available,
 		 * just use it.
@@ -2337,7 +2337,7 @@ static int tcp_mtu_probe(struct sock *sk)
 	if (likely(!icsk->icsk_mtup.enabled ||
 		   icsk->icsk_mtup.probe_size ||
 		   inet_csk(sk)->icsk_ca_state != TCP_CA_Open ||
-		   tp->snd_cwnd < 11 ||
+		   tcp_snd_cwnd(tp) < 11 ||
 		   tp->rx_opt.num_sacks || tp->rx_opt.dsack))
 		return -1;
 
@@ -2373,7 +2373,7 @@ static int tcp_mtu_probe(struct sock *sk)
 		return 0;
 
 	/* Do we need to wait to drain cwnd? With none in flight, don't stall */
-	if (tcp_packets_in_flight(tp) + 2 > tp->snd_cwnd) {
+	if (tcp_packets_in_flight(tp) + 2 > tcp_snd_cwnd(tp)) {
 		if (!tcp_packets_in_flight(tp))
 			return -1;
 		else
@@ -2442,7 +2442,7 @@ static int tcp_mtu_probe(struct sock *sk)
 	if (!tcp_transmit_skb(sk, nskb, 1, GFP_ATOMIC)) {
 		/* Decrement cwnd here because we are sending
 		 * effectively two packets. */
-		tp->snd_cwnd--;
+		tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) - 1);
 		tcp_event_new_data_sent(sk, nskb);
 
 		icsk->icsk_mtup.probe_size = tcp_mss_to_mtu(sk, nskb->len);
@@ -2699,7 +2699,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 	else
 		tcp_chrono_stop(sk, TCP_CHRONO_RWND_LIMITED);
 
-	is_cwnd_limited |= (tcp_packets_in_flight(tp) >= tp->snd_cwnd);
+	is_cwnd_limited |= (tcp_packets_in_flight(tp) >= tcp_snd_cwnd(tp));
 	if (likely(sent_pkts || is_cwnd_limited))
 		tcp_cwnd_validate(sk, is_cwnd_limited);
 
@@ -2809,7 +2809,7 @@ void tcp_send_loss_probe(struct sock *sk)
 	if (unlikely(!skb)) {
 		WARN_ONCE(tp->packets_out,
 			  "invalid inflight: %u state %u cwnd %u mss %d\n",
-			  tp->packets_out, sk->sk_state, tp->snd_cwnd, mss);
+			  tp->packets_out, sk->sk_state, tcp_snd_cwnd(tp), mss);
 		inet_csk(sk)->icsk_pending = 0;
 		return;
 	}
@@ -3293,7 +3293,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk)
 		if (!hole)
 			tp->retransmit_skb_hint = skb;
 
-		segs = tp->snd_cwnd - tcp_packets_in_flight(tp);
+		segs = tcp_snd_cwnd(tp) - tcp_packets_in_flight(tp);
 		if (segs <= 0)
 			break;
 		sacked = TCP_SKB_CB(skb)->sacked;
net/ipv4/tcp_rate.c +1 -1
index 9a8e014d9b5b..a8f6d9d06f2e 100644
@@ -200,7 +200,7 @@ void tcp_rate_check_app_limited(struct sock *sk)
 	    /* Nothing in sending host's qdisc queues or NIC tx queue. */
 	    sk_wmem_alloc_get(sk) < SKB_TRUESIZE(1) &&
 	    /* We are not limited by CWND. */
-	    tcp_packets_in_flight(tp) < tp->snd_cwnd &&
+	    tcp_packets_in_flight(tp) < tcp_snd_cwnd(tp) &&
 	    /* All lost packets have been retransmitted. */
 	    tp->lost_out <= tp->retrans_out)
 		tp->app_limited =
net/ipv4/tcp_scalable.c +2 -2
index 5842081bc8a2..862b96248a92 100644
@@ -27,7 +27,7 @@ static void tcp_scalable_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		if (!acked)
 			return;
 	}
-	tcp_cong_avoid_ai(tp, min(tp->snd_cwnd, TCP_SCALABLE_AI_CNT),
+	tcp_cong_avoid_ai(tp, min(tcp_snd_cwnd(tp), TCP_SCALABLE_AI_CNT),
 			  acked);
 }
 
@@ -35,7 +35,7 @@ static u32 tcp_scalable_ssthresh(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 
-	return max(tp->snd_cwnd - (tp->snd_cwnd>>TCP_SCALABLE_MD_SCALE), 2U);
+	return max(tcp_snd_cwnd(tp) - (tcp_snd_cwnd(tp)>>TCP_SCALABLE_MD_SCALE), 2U);
 }
 
 static struct tcp_congestion_ops tcp_scalable __read_mostly = {
net/ipv4/tcp_vegas.c +11 -10
index c8003c8aad2c..786848ad37ea 100644
@@ -159,7 +159,7 @@ EXPORT_SYMBOL_GPL(tcp_vegas_cwnd_event);
 
 static inline u32 tcp_vegas_ssthresh(struct tcp_sock *tp)
 {
-	return  min(tp->snd_ssthresh, tp->snd_cwnd);
+	return  min(tp->snd_ssthresh, tcp_snd_cwnd(tp));
 }
 
 static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked)
@@ -217,14 +217,14 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 			 * This is:
 			 *     (actual rate in segments) * baseRTT
 			 */
-			target_cwnd = (u64)tp->snd_cwnd * vegas->baseRTT;
+			target_cwnd = (u64)tcp_snd_cwnd(tp) * vegas->baseRTT;
 			do_div(target_cwnd, rtt);
 
 			/* Calculate the difference between the window we had,
 			 * and the window we would like to have. This quantity
 			 * is the "Diff" from the Arizona Vegas papers.
 			 */
-			diff = tp->snd_cwnd * (rtt-vegas->baseRTT) / vegas->baseRTT;
+			diff = tcp_snd_cwnd(tp) * (rtt-vegas->baseRTT) / vegas->baseRTT;
 
 			if (diff > gamma && tcp_in_slow_start(tp)) {
 				/* Going too fast. Time to slow down
@@ -238,7 +238,8 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 				 * truncation robs us of full link
 				 * utilization.
 				 */
-				tp->snd_cwnd = min(tp->snd_cwnd, (u32)target_cwnd+1);
+				tcp_snd_cwnd_set(tp, min(tcp_snd_cwnd(tp),
+							 (u32)target_cwnd + 1));
 				tp->snd_ssthresh = tcp_vegas_ssthresh(tp);
 
 			} else if (tcp_in_slow_start(tp)) {
@@ -254,14 +255,14 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 					/* The old window was too fast, so
 					 * we slow down.
 					 */
-					tp->snd_cwnd--;
+					tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) - 1);
 					tp->snd_ssthresh
 						= tcp_vegas_ssthresh(tp);
 				} else if (diff < alpha) {
 					/* We don't have enough extra packets
 					 * in the network, so speed up.
 					 */
-					tp->snd_cwnd++;
+					tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 				} else {
 					/* Sending just as fast as we
 					 * should be.
@@ -269,10 +270,10 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 				}
 			}
 
-			if (tp->snd_cwnd < 2)
-				tp->snd_cwnd = 2;
-			else if (tp->snd_cwnd > tp->snd_cwnd_clamp)
-				tp->snd_cwnd = tp->snd_cwnd_clamp;
+			if (tcp_snd_cwnd(tp) < 2)
+				tcp_snd_cwnd_set(tp, 2);
+			else if (tcp_snd_cwnd(tp) > tp->snd_cwnd_clamp)
+				tcp_snd_cwnd_set(tp, tp->snd_cwnd_clamp);
 
 			tp->snd_ssthresh = tcp_current_ssthresh(sk);
 		}
net/ipv4/tcp_veno.c +12 -12
index cd50a61c9976..366ff6f214b2 100644
@@ -146,11 +146,11 @@ static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 
 		rtt = veno->minrtt;
 
-		target_cwnd = (u64)tp->snd_cwnd * veno->basertt;
+		target_cwnd = (u64)tcp_snd_cwnd(tp) * veno->basertt;
 		target_cwnd <<= V_PARAM_SHIFT;
 		do_div(target_cwnd, rtt);
 
-		veno->diff = (tp->snd_cwnd << V_PARAM_SHIFT) - target_cwnd;
+		veno->diff = (tcp_snd_cwnd(tp) << V_PARAM_SHIFT) - target_cwnd;
 
 		if (tcp_in_slow_start(tp)) {
 			/* Slow start. */
@@ -164,15 +164,15 @@ static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 			/* In the "non-congestive state", increase cwnd
 			 * every rtt.
 			 */
-			tcp_cong_avoid_ai(tp, tp->snd_cwnd, acked);
+			tcp_cong_avoid_ai(tp, tcp_snd_cwnd(tp), acked);
 		} else {
 			/* In the "congestive state", increase cwnd
 			 * every other rtt.
 			 */
-			if (tp->snd_cwnd_cnt >= tp->snd_cwnd) {
+			if (tp->snd_cwnd_cnt >= tcp_snd_cwnd(tp)) {
 				if (veno->inc &&
-				    tp->snd_cwnd < tp->snd_cwnd_clamp) {
-					tp->snd_cwnd++;
+				    tcp_snd_cwnd(tp) < tp->snd_cwnd_clamp) {
+					tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) + 1);
 					veno->inc = 0;
 				} else
 					veno->inc = 1;
@@ -181,10 +181,10 @@ static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 				tp->snd_cwnd_cnt += acked;
 		}
 done:
-		if (tp->snd_cwnd < 2)
-			tp->snd_cwnd = 2;
-		else if (tp->snd_cwnd > tp->snd_cwnd_clamp)
-			tp->snd_cwnd = tp->snd_cwnd_clamp;
+		if (tcp_snd_cwnd(tp) < 2)
+			tcp_snd_cwnd_set(tp, 2);
+		else if (tcp_snd_cwnd(tp) > tp->snd_cwnd_clamp)
+			tcp_snd_cwnd_set(tp, tp->snd_cwnd_clamp);
 	}
 	/* Wipe the slate clean for the next rtt. */
 	/* veno->cntrtt = 0; */
@@ -199,10 +199,10 @@ static u32 tcp_veno_ssthresh(struct sock *sk)
 
 	if (veno->diff < beta)
 		/* in "non-congestive state", cut cwnd by 1/5 */
-		return max(tp->snd_cwnd * 4 / 5, 2U);
+		return max(tcp_snd_cwnd(tp) * 4 / 5, 2U);
 	else
 		/* in "congestive state", cut cwnd by 1/2 */
-		return max(tp->snd_cwnd >> 1U, 2U);
+		return max(tcp_snd_cwnd(tp) >> 1U, 2U);
 }
 
 static struct tcp_congestion_ops tcp_veno __read_mostly = {
net/ipv4/tcp_westwood.c +2 -1
index b2e05c4cea00..c6e97141eef2 100644
@@ -244,7 +244,8 @@ static void tcp_westwood_event(struct sock *sk, enum tcp_ca_event event)
 
 	switch (event) {
 	case CA_EVENT_COMPLETE_CWR:
-		tp->snd_cwnd = tp->snd_ssthresh = tcp_westwood_bw_rttmin(sk);
+		tp->snd_ssthresh = tcp_westwood_bw_rttmin(sk);
+		tcp_snd_cwnd_set(tp, tp->snd_ssthresh);
 		break;
 	case CA_EVENT_LOSS:
 		tp->snd_ssthresh = tcp_westwood_bw_rttmin(sk);
net/ipv4/tcp_yeah.c +15 -15
index 07c4c93b9fdb..18b07ff5d20e 100644
@@ -71,11 +71,11 @@ static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 
 	if (!yeah->doing_reno_now) {
 		/* Scalable */
-		tcp_cong_avoid_ai(tp, min(tp->snd_cwnd, TCP_SCALABLE_AI_CNT),
+		tcp_cong_avoid_ai(tp, min(tcp_snd_cwnd(tp), TCP_SCALABLE_AI_CNT),
 				  acked);
 	} else {
 		/* Reno */
-		tcp_cong_avoid_ai(tp, tp->snd_cwnd, acked);
+		tcp_cong_avoid_ai(tp, tcp_snd_cwnd(tp), acked);
 	}
 
 	/* The key players are v_vegas.beg_snd_una and v_beg_snd_nxt.
@@ -130,7 +130,7 @@ static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 			/* Compute excess number of packets above bandwidth
 			 * Avoid doing full 64 bit divide.
 			 */
-			bw = tp->snd_cwnd;
+			bw = tcp_snd_cwnd(tp);
 			bw *= rtt - yeah->vegas.baseRTT;
 			do_div(bw, rtt);
 			queue = bw;
@@ -138,20 +138,20 @@ static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 			if (queue > TCP_YEAH_ALPHA ||
 			    rtt - yeah->vegas.baseRTT > (yeah->vegas.baseRTT / TCP_YEAH_PHY)) {
 				if (queue > TCP_YEAH_ALPHA &&
-				    tp->snd_cwnd > yeah->reno_count) {
+				    tcp_snd_cwnd(tp) > yeah->reno_count) {
 					u32 reduction = min(queue / TCP_YEAH_GAMMA ,
-							    tp->snd_cwnd >> TCP_YEAH_EPSILON);
+							    tcp_snd_cwnd(tp) >> TCP_YEAH_EPSILON);
 
-					tp->snd_cwnd -= reduction;
+					tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) - reduction);
 
-					tp->snd_cwnd = max(tp->snd_cwnd,
-							   yeah->reno_count);
+					tcp_snd_cwnd_set(tp, max(tcp_snd_cwnd(tp),
+								 yeah->reno_count));
 
-					tp->snd_ssthresh = tp->snd_cwnd;
+					tp->snd_ssthresh = tcp_snd_cwnd(tp);
 				}
 
 				if (yeah->reno_count <= 2)
-					yeah->reno_count = max(tp->snd_cwnd>>1, 2U);
+					yeah->reno_count = max(tcp_snd_cwnd(tp)>>1, 2U);
 				else
 					yeah->reno_count++;
 
@@ -176,7 +176,7 @@ static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 		 */
 		yeah->vegas.beg_snd_una  = yeah->vegas.beg_snd_nxt;
 		yeah->vegas.beg_snd_nxt  = tp->snd_nxt;
-		yeah->vegas.beg_snd_cwnd = tp->snd_cwnd;
+		yeah->vegas.beg_snd_cwnd = tcp_snd_cwnd(tp);
 
 		/* Wipe the slate clean for the next RTT. */
 		yeah->vegas.cntRTT = 0;
@@ -193,16 +193,16 @@ static u32 tcp_yeah_ssthresh(struct sock *sk)
 	if (yeah->doing_reno_now < TCP_YEAH_RHO) {
 		reduction = yeah->lastQ;
 
-		reduction = min(reduction, max(tp->snd_cwnd>>1, 2U));
+		reduction = min(reduction, max(tcp_snd_cwnd(tp)>>1, 2U));
 
-		reduction = max(reduction, tp->snd_cwnd >> TCP_YEAH_DELTA);
+		reduction = max(reduction, tcp_snd_cwnd(tp) >> TCP_YEAH_DELTA);
 	} else
-		reduction = max(tp->snd_cwnd>>1, 2U);
+		reduction = max(tcp_snd_cwnd(tp)>>1, 2U);
 
 	yeah->fast_count = 0;
 	yeah->reno_count = max(yeah->reno_count>>1, 2U);
 
-	return max_t(int, tp->snd_cwnd - reduction, 2);
+	return max_t(int, tcp_snd_cwnd(tp) - reduction, 2);
 }
 
 static struct tcp_congestion_ops tcp_yeah __read_mostly = {
net/ipv6/tcp_ipv6.c +1 -1
index 075ee8a2df3b..29a4fc92580e 100644
@@ -2074,7 +2074,7 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i)
 		   jiffies_to_clock_t(icsk->icsk_rto),
 		   jiffies_to_clock_t(icsk->icsk_ack.ato),
 		   (icsk->icsk_ack.quick << 1) | inet_csk_in_pingpong_mode(sp),
-		   tp->snd_cwnd,
+		   tcp_snd_cwnd(tp),
 		   state == TCP_LISTEN ?
 			fastopenq->max_qlen :
 			(tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh)
Signature · 1 line
2.35.1
author maintainer
#100 lore raw

PATCH 100/298 nfp: only report pause frame configuration for physical device

In reply to Greg Kroah-Hartman

From: Yu Xiao <[email protected]>

[ Upstream commit 0649e4d63420ebc8cbebef3e9d39e12ffc5eb9fa ]

Only report pause frame configuration for physical device. Logical port of both PCI PF and PCI VF do not support it.

Fixes 9fdc5d85a8fe ("nfp: update ethtool reporting of pauseframe control")
---
 drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c +2 -2
index e0c27471bcdb..5e2631aafdb6 100644
@@ -287,8 +287,6 @@ nfp_net_get_link_ksettings(struct net_device *netdev,
 
 	/* Init to unknowns */
 	ethtool_link_ksettings_add_link_mode(cmd, supported, FIBRE);
-	ethtool_link_ksettings_add_link_mode(cmd, supported, Pause);
-	ethtool_link_ksettings_add_link_mode(cmd, advertising, Pause);
 	cmd->base.port = PORT_OTHER;
 	cmd->base.speed = SPEED_UNKNOWN;
 	cmd->base.duplex = DUPLEX_UNKNOWN;
@@ -296,6 +294,8 @@ nfp_net_get_link_ksettings(struct net_device *netdev,
 	port = nfp_port_from_netdev(netdev);
 	eth_port = nfp_port_get_eth_port(port);
 	if (eth_port) {
+		ethtool_link_ksettings_add_link_mode(cmd, supported, Pause);
+		ethtool_link_ksettings_add_link_mode(cmd, advertising, Pause);
 		cmd->base.autoneg = eth_port->aneg != NFP_ANEG_DISABLED ?
 			AUTONEG_ENABLE : AUTONEG_DISABLE;
 		nfp_net_set_fec_link_mode(eth_port, cmd);
Signature · 1 line
2.35.1
author maintainer
#101 lore raw

PATCH 101/298 block: use bio_queue_enter instead of blk_queue_enter in bio_poll

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit ebd076bf7d5deef488ec7ebc3fdbf781eafae269 ]

We want to have a valid live gendisk to call ->poll and not just a request_queue, so call the right helper.

Fixes 3e08773c3841 ("block: switch polling to be bio based")
---
 block/blk-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/blk-core.c +1 -1
index 779b4a1f66ac..45d750eb2628 100644
@@ -992,7 +992,7 @@ int bio_poll(struct bio *bio, struct io_comp_batch *iob, unsigned int flags)
 	if (current->plug)
 		blk_flush_plug(current->plug, false);
 
-	if (blk_queue_enter(q, BLK_MQ_REQ_NOWAIT))
+	if (bio_queue_enter(bio))
 		return 0;
 	if (WARN_ON_ONCE(!queue_is_mq(q)))
 		ret = 0;	/* not yet implemented, should not happen */
Signature · 1 line
2.35.1
author maintainer
#102 lore raw

PATCH 102/298 sfc: fix considering that all channels have TX queues

In reply to Greg Kroah-Hartman

From: Martin Habets <[email protected]>

[ Upstream commit 2e102b53f8a778f872dc137f4c7ac548705817aa ]

Normally, all channels have RX and TX queues, but this is not true if modparam efx_separate_tx_channels=1 is used. In that cases, some channels only have RX queues and others only TX queues (or more preciselly, they have them allocated, but not initialized).

Fix efx_channel_has_tx_queues to return the correct value for this case too.

Messages shown at probe time before the fix:
sfc 0000:03:00.0 ens6f0np0: MC command 0x82 inlen 544 failed rc=-22 (raw=0) arg=0

 ------------[ cut here ]------------
 netdevice: ens6f0np0: failed to initialise TXQ -1
 WARNING: CPU: 1 PID: 626 at drivers/net/ethernet/sfc/ef10.c:2393 efx_ef10_tx_init+0x201/0x300 [sfc]
 [...] stripped
 RIP: 0010:efx_ef10_tx_init+0x201/0x300 [sfc]
 [...] stripped
 Call Trace:
  efx_init_tx_queue+0xaa/0xf0 [sfc]
  efx_start_channels+0x49/0x120 [sfc]
  efx_start_all+0x1f8/0x430 [sfc]
  efx_net_open+0x5a/0xe0 [sfc]
  __dev_open+0xd0/0x190
  __dev_change_flags+0x1b3/0x220
  dev_change_flags+0x21/0x60
 [...] stripped

Messages shown at remove time before the fix:
sfc 0000:03:00.0 ens6f0np0: failed to flush 10 queues sfc 0000:03:00.0 ens6f0np0: failed to flush queues

Fixes 8700aff08984 ("sfc: fix channel allocation with brute force")
---
 drivers/net/ethernet/sfc/net_driver.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/sfc/net_driver.h +1 -1
index cc15ee8812d9..8a9eedec177a 100644
@@ -1533,7 +1533,7 @@ static inline bool efx_channel_is_xdp_tx(struct efx_channel *channel)
 
 static inline bool efx_channel_has_tx_queues(struct efx_channel *channel)
 {
-	return true;
+	return channel && channel->channel >= channel->efx->tx_channel_offset;
 }
 
 static inline unsigned int efx_channel_num_tx_queues(struct efx_channel *channel)
Signature · 1 line
2.35.1
author maintainer
#103 lore raw

PATCH 103/298 sfc: fix wrong tx channel offset with efx_separate_tx_channels

In reply to Greg Kroah-Hartman

From: Íñigo Huguet <[email protected]>

[ Upstream commit c308dfd1b43ef0d4c3e57b741bb3462eb7a7f4a2 ]

tx_channel_offset is calculated in efx_allocate_msix_channels, but it is also calculated again in efx_set_channels because it was originally done there, and when efx_allocate_msix_channels was introduced it was forgotten to be removed from efx_set_channels.

Moreover, the old calculation is wrong when using efx_separate_tx_channels because now we can have XDP channels after the TX channels, so n_channels - n_tx_channels doesn't point to the first TX channel.

Remove the old calculation from efx_set_channels, and add the initialization of this variable if MSI or legacy interrupts are used, next to the initialization of the rest of the related variables, where it was missing.

Fixes 3990a8fffbda ("sfc: allocate channels for XDP tx queues")
---
 drivers/net/ethernet/sfc/efx_channels.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/net/ethernet/sfc/efx_channels.c +2 -4
index eec0db76d888..8ab9358a1c3d 100644
@@ -309,6 +309,7 @@ int efx_probe_interrupts(struct efx_nic *efx)
 		efx->n_channels = 1;
 		efx->n_rx_channels = 1;
 		efx->n_tx_channels = 1;
+		efx->tx_channel_offset = 0;
 		efx->n_xdp_channels = 0;
 		efx->xdp_channel_offset = efx->n_channels;
 		rc = pci_enable_msi(efx->pci_dev);
@@ -329,6 +330,7 @@ int efx_probe_interrupts(struct efx_nic *efx)
 		efx->n_channels = 1 + (efx_separate_tx_channels ? 1 : 0);
 		efx->n_rx_channels = 1;
 		efx->n_tx_channels = 1;
+		efx->tx_channel_offset = 1;
 		efx->n_xdp_channels = 0;
 		efx->xdp_channel_offset = efx->n_channels;
 		efx->legacy_irq = efx->pci_dev->irq;
@@ -957,10 +959,6 @@ int efx_set_channels(struct efx_nic *efx)
 	struct efx_channel *channel;
 	int rc;
 
-	efx->tx_channel_offset =
-		efx_separate_tx_channels ?
-		efx->n_channels - efx->n_tx_channels : 0;
-
 	if (efx->xdp_tx_queue_count) {
 		EFX_WARN_ON_PARANOID(efx->xdp_tx_queues);
 
Signature · 1 line
2.35.1
author maintainer
#104 lore raw

PATCH 104/298 block: make bioset_exit() fully resilient against being called twice

In reply to Greg Kroah-Hartman

From: Jens Axboe <[email protected]>

[ Upstream commit 605f7415ecfb426610195dd6c7577b30592b3369 ]

Most of bioset_exit() is fine being called twice, as it clears the various allocations etc when they are freed. The exception is bio_alloc_cache_destroy(), which does not clear ->cache when it has freed it.

This isn't necessarily a bug, but can be if buggy users does call the exit path more then once, or with just a memset() bioset which has never been initialized. dm appears to be one such user.

---
 block/bio.c | 1 +
 1 file changed, 1 insertion(+)
block/bio.c +1 -0
index 738fea03edbf..dc6940621d7d 100644
@@ -668,6 +668,7 @@ static void bio_alloc_cache_destroy(struct bio_set *bs)
 		bio_alloc_cache_prune(cache, -1U);
 	}
 	free_percpu(bs->cache);
+	bs->cache = NULL;
 }
 
 /**
Signature · 1 line
2.35.1
author maintainer
#105 lore raw

PATCH 105/298 blk-mq: do not update io_ticks with passthrough requests

In reply to Greg Kroah-Hartman

From: Haisu Wang <[email protected]>

[ Upstream commit b81c14ca14b631aa1abae32fb5ae75b5e9251012 ]

Flush or passthrough requests are not accounted as normal IO in completion. To reflect iostat for slow IO, io_ticks is updated when stat show called based on inflight numbers.
It may cause inconsistent io_ticks calculation result.

So do not account non-passthrough request when check inflight.

---
 block/blk-mq.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
block/blk-mq.c +2 -1
index 6afe0cd128ac..f18e1c9c3f4a 100644
@@ -132,7 +132,8 @@ static bool blk_mq_check_inflight(struct request *rq, void *priv,
 {
 	struct mq_inflight *mi = priv;
 
-	if ((!mi->part->bd_partno || rq->part == mi->part) &&
+	if (rq->part && blk_do_io_stat(rq) &&
+	    (!mi->part->bd_partno || rq->part == mi->part) &&
 	    blk_mq_rq_state(rq) == MQ_RQ_IN_FLIGHT)
 		mi->inflight[rq_data_dir(rq)]++;
 
Signature · 1 line
2.35.1
author maintainer
#106 lore raw

PATCH 106/298 vdpa: Fix error logic in vdpa_nl_cmd_dev_get_doit

In reply to Greg Kroah-Hartman

From: Eli Cohen <[email protected]>

[ Upstream commit 7a6691f1f89784f775fa0c54be57533445726068 ]

In vdpa_nl_cmd_dev_get_doit(), if the call to genlmsg_reply() fails we must not call nlmsg_free() since this is done inside genlmsg_reply().

Fix it.

Fixes bc0d90ee021f ("vdpa: Enable user to query vdpa device info")

Message-Id: <[email protected]>

---
 drivers/vdpa/vdpa.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
drivers/vdpa/vdpa.c +9 -4
index 1ea525433a5c..4e7351110e43 100644
@@ -756,14 +756,19 @@ static int vdpa_nl_cmd_dev_get_doit(struct sk_buff *skb, struct genl_info *info)
 		goto mdev_err;
 	}
 	err = vdpa_dev_fill(vdev, msg, info->snd_portid, info->snd_seq, 0, info->extack);
-	if (!err)
-		err = genlmsg_reply(msg, info);
+	if (err)
+		goto mdev_err;
+
+	err = genlmsg_reply(msg, info);
+	put_device(dev);
+	mutex_unlock(&vdpa_dev_mutex);
+	return err;
+
 mdev_err:
 	put_device(dev);
 err:
 	mutex_unlock(&vdpa_dev_mutex);
-	if (err)
-		nlmsg_free(msg);
+	nlmsg_free(msg);
 	return err;
 }
 
Signature · 1 line
2.35.1
author maintainer
#107 lore raw

PATCH 107/298 virtio: pci: Fix an error handling path in vp_modern_probe()

In reply to Greg Kroah-Hartman

From: Christophe JAILLET <[email protected]>

[ Upstream commit 7a836a2aba09479c8e71fa43249eecc4af945f61 ]

If an error occurs after a successful pci_request_selected_regions() call, it should be undone by a corresponding pci_release_selected_regions() call, as already done in vp_modern_remove().

Fixes fd502729fbbf ("virtio-pci: introduce modern device module")

Message-Id: <237109725aad2c3c03d14549f777b1927c84b045.1648977064.git.christophe.jaillet@wanadoo.fr>

---
 drivers/virtio/virtio_pci_modern_dev.c | 1 +
 1 file changed, 1 insertion(+)
drivers/virtio/virtio_pci_modern_dev.c +1 -0
index e8b3ff2b9fbc..6f6187fe8893 100644
@@ -340,6 +340,7 @@ int vp_modern_probe(struct virtio_pci_modern_device *mdev)
 err_map_isr:
 	pci_iounmap(pci_dev, mdev->common);
 err_map_common:
+	pci_release_selected_regions(pci_dev, mdev->modern_bars);
 	return err;
 }
 EXPORT_SYMBOL_GPL(vp_modern_probe);
Signature · 1 line
2.35.1
author maintainer
#108 lore raw

PATCH 108/298 net/mlx5: Dont use already freed action pointer

In reply to Greg Kroah-Hartman

From: Leon Romanovsky <[email protected]>

[ Upstream commit 80b2bd737d0e833e6a2b77e482e5a714a79c86a4 ]

The call to mlx5dr_action_destroy() releases "action" memory. That pointer is set to miss_action later and generates the following smatch error:

drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c:53 set_miss_action()
warn: 'action' was already freed.

Make sure that the pointer is always valid by setting NULL after destroy.

Fixes 6a48faeeca10 ("net/mlx5: Add direct rule fs_cmd implementation")
---
 drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c +4 -5
index 05393fe11132..caeaa3c29353 100644
@@ -44,11 +44,10 @@ static int set_miss_action(struct mlx5_flow_root_namespace *ns,
 	err = mlx5dr_table_set_miss_action(ft->fs_dr_table.dr_table, action);
 	if (err && action) {
 		err = mlx5dr_action_destroy(action);
-		if (err) {
-			action = NULL;
-			mlx5_core_err(ns->dev, "Failed to destroy action (%d)\n",
-				      err);
-		}
+		if (err)
+			mlx5_core_err(ns->dev,
+				      "Failed to destroy action (%d)\n", err);
+		action = NULL;
 	}
 	ft->fs_dr_table.miss_action = action;
 	if (old_miss_action) {
Signature · 1 line
2.35.1
author maintainer
#109 lore raw

PATCH 109/298 net/mlx5e: TC NIC mode, fix tc chains miss table

In reply to Greg Kroah-Hartman

From: Maor Dickman <[email protected]>

[ Upstream commit 66cb64e292d21588bdb831f08a7ec0ff04d6380d ]

The cited commit changed promisc table to be created on demand with the highest priority in the NIC table replacing the vlan table, this caused tc NIC tables miss flow to skip the prmoisc table because it use vlan table as miss table.

OVS offload in NIC mode use promisc by default so any unicast packet which will be handled by tc NIC tables miss flow will skip the promisc rule and will be dropped.

Fix this by adding new empty table in new tc level with low priority and point the nic tc chain miss to it, the new table is managed so it will point to vlan table if promisc is disabled and to promisc table if enabled.

Fixes 1c46d7409f30 ("net/mlx5e: Optimize promiscuous mode")
---
 .../net/ethernet/mellanox/mlx5/core/en/fs.h   |  2 +
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 38 ++++++++++++++++++-
 .../net/ethernet/mellanox/mlx5/core/fs_core.c |  2 +-
 3 files changed, 39 insertions(+), 3 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/fs.h +2 -0
index 678ffbb48a25..e3e8c1c3ff24 100644
@@ -12,6 +12,7 @@ struct mlx5e_post_act;
 enum {
 	MLX5E_TC_FT_LEVEL = 0,
 	MLX5E_TC_TTC_FT_LEVEL,
+	MLX5E_TC_MISS_LEVEL,
 };
 
 struct mlx5e_tc_table {
@@ -20,6 +21,7 @@ struct mlx5e_tc_table {
 	 */
 	struct mutex			t_lock;
 	struct mlx5_flow_table		*t;
+	struct mlx5_flow_table		*miss_t;
 	struct mlx5_fs_chains           *chains;
 	struct mlx5e_post_act		*post_act;
 
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +36 -2
index e0f45cef97c3..deff6698f395 100644
@@ -4284,6 +4284,33 @@ static int mlx5e_tc_nic_get_ft_size(struct mlx5_core_dev *dev)
 	return tc_tbl_size;
 }
 
+static int mlx5e_tc_nic_create_miss_table(struct mlx5e_priv *priv)
+{
+	struct mlx5_flow_table **ft = &priv->fs.tc.miss_t;
+	struct mlx5_flow_table_attr ft_attr = {};
+	struct mlx5_flow_namespace *ns;
+	int err = 0;
+
+	ft_attr.max_fte = 1;
+	ft_attr.autogroup.max_num_groups = 1;
+	ft_attr.level = MLX5E_TC_MISS_LEVEL;
+	ft_attr.prio = 0;
+	ns = mlx5_get_flow_namespace(priv->mdev, MLX5_FLOW_NAMESPACE_KERNEL);
+
+	*ft = mlx5_create_auto_grouped_flow_table(ns, &ft_attr);
+	if (IS_ERR(*ft)) {
+		err = PTR_ERR(*ft);
+		netdev_err(priv->netdev, "failed to create tc nic miss table err=%d\n", err);
+	}
+
+	return err;
+}
+
+static void mlx5e_tc_nic_destroy_miss_table(struct mlx5e_priv *priv)
+{
+	mlx5_destroy_flow_table(priv->fs.tc.miss_t);
+}
+
 int mlx5e_tc_nic_init(struct mlx5e_priv *priv)
 {
 	struct mlx5e_tc_table *tc = &priv->fs.tc;
@@ -4316,19 +4343,23 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv)
 	}
 	tc->mapping = chains_mapping;
 
+	err = mlx5e_tc_nic_create_miss_table(priv);
+	if (err)
+		goto err_chains;
+
 	if (MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ignore_flow_level))
 		attr.flags = MLX5_CHAINS_AND_PRIOS_SUPPORTED |
 			MLX5_CHAINS_IGNORE_FLOW_LEVEL_SUPPORTED;
 	attr.ns = MLX5_FLOW_NAMESPACE_KERNEL;
 	attr.max_ft_sz = mlx5e_tc_nic_get_ft_size(dev);
 	attr.max_grp_num = MLX5E_TC_TABLE_NUM_GROUPS;
-	attr.default_ft = mlx5e_vlan_get_flowtable(priv->fs.vlan);
+	attr.default_ft = priv->fs.tc.miss_t;
 	attr.mapping = chains_mapping;
 
 	tc->chains = mlx5_chains_create(dev, &attr);
 	if (IS_ERR(tc->chains)) {
 		err = PTR_ERR(tc->chains);
-		goto err_chains;
+		goto err_miss;
 	}
 
 	tc->post_act = mlx5e_tc_post_act_init(priv, tc->chains, MLX5_FLOW_NAMESPACE_KERNEL);
@@ -4351,6 +4382,8 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv)
 	mlx5_tc_ct_clean(tc->ct);
 	mlx5e_tc_post_act_destroy(tc->post_act);
 	mlx5_chains_destroy(tc->chains);
+err_miss:
+	mlx5e_tc_nic_destroy_miss_table(priv);
 err_chains:
 	mapping_destroy(chains_mapping);
 err_mapping:
@@ -4391,6 +4424,7 @@ void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv)
 	mlx5e_tc_post_act_destroy(tc->post_act);
 	mapping_destroy(tc->mapping);
 	mlx5_chains_destroy(tc->chains);
+	mlx5e_tc_nic_destroy_miss_table(priv);
 }
 
 int mlx5e_tc_esw_init(struct rhashtable *tc_ht)
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +1 -1
index 298c614c631b..bd52b0b4eefa 100644
@@ -116,7 +116,7 @@
 #define KERNEL_MIN_LEVEL (KERNEL_NIC_PRIO_NUM_LEVELS + 1)
 
 #define KERNEL_NIC_TC_NUM_PRIOS  1
-#define KERNEL_NIC_TC_NUM_LEVELS 2
+#define KERNEL_NIC_TC_NUM_LEVELS 3
 
 #define ANCHOR_NUM_LEVELS 1
 #define ANCHOR_NUM_PRIOS 1
Signature · 1 line
2.35.1
author maintainer
#110 lore raw

PATCH 110/298 net/mlx5: CT: Fix header-rewrite re-use for tupels

In reply to Greg Kroah-Hartman

From: Paul Blakey <[email protected]>

[ Upstream commit 1f2856cde64baa78475e6d3c601fb7b7f693a161 ]

Tuple entries that don't have nat configured for them which are added to the ct nat table will always create a new modify header, as we don't check for possible re-use on them. The same for tuples that have nat configured for them but are added to ct table.

Fix the above by only avoiding wasteful re-use lookup for actually natted entries in ct nat table.

Fixes 7fac5c2eced3 ("net/mlx5: CT: Avoid reusing modify header context for natted entries")
---
 .../ethernet/mellanox/mlx5/core/en/tc_ct.c    | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +11 -8
index 9028e9958c72..cf9d48d934ef 100644
@@ -692,7 +692,7 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv,
 				struct mlx5_flow_attr *attr,
 				struct flow_rule *flow_rule,
 				struct mlx5e_mod_hdr_handle **mh,
-				u8 zone_restore_id, bool nat)
+				u8 zone_restore_id, bool nat_table, bool has_nat)
 {
 	DECLARE_MOD_HDR_ACTS_ACTIONS(actions_arr, MLX5_CT_MIN_MOD_ACTS);
 	DECLARE_MOD_HDR_ACTS(mod_acts, actions_arr);
@@ -708,11 +708,12 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv,
 				     &attr->ct_attr.ct_labels_id);
 	if (err)
 		return -EOPNOTSUPP;
-	if (nat) {
-		err = mlx5_tc_ct_entry_create_nat(ct_priv, flow_rule,
-						  &mod_acts);
-		if (err)
-			goto err_mapping;
+	if (nat_table) {
+		if (has_nat) {
+			err = mlx5_tc_ct_entry_create_nat(ct_priv, flow_rule, &mod_acts);
+			if (err)
+				goto err_mapping;
+		}
 
 		ct_state |= MLX5_CT_STATE_NAT_BIT;
 	}
@@ -727,7 +728,7 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv,
 	if (err)
 		goto err_mapping;
 
-	if (nat) {
+	if (nat_table && has_nat) {
 		attr->modify_hdr = mlx5_modify_header_alloc(ct_priv->dev, ct_priv->ns_type,
 							    mod_acts.num_actions,
 							    mod_acts.actions);
@@ -795,7 +796,9 @@ mlx5_tc_ct_entry_add_rule(struct mlx5_tc_ct_priv *ct_priv,
 
 	err = mlx5_tc_ct_entry_create_mod_hdr(ct_priv, attr, flow_rule,
 					      &zone_rule->mh,
-					      zone_restore_id, nat);
+					      zone_restore_id,
+					      nat,
+					      mlx5_tc_ct_entry_has_nat(entry));
 	if (err) {
 		ct_dbg("Failed to create ct entry mod hdr");
 		goto err_mod_hdr;
Signature · 1 line
2.35.1
author maintainer
#111 lore raw

PATCH 111/298 net/mlx5e: Disable softirq in mlx5e_activate_rq to avoid race condition

In reply to Greg Kroah-Hartman

From: Maxim Mikityanskiy <[email protected]>

[ Upstream commit 2e642afb61b24401a7ec819d27ddcd69c7c29784 ]

When the driver activates the channels, it assumes NAPI isn't running yet. mlx5e_activate_rq posts a NOP WQE to ICOSQ to trigger a hardware interrupt and start NAPI, which will run mlx5e_alloc_rx_mpwqe and post UMR WQEs to ICOSQ to be able to receive packets with striding RQ.

Unfortunately, a race condition is possible if NAPI is triggered by something else (for example, TX) at a bad timing, before mlx5e_activate_rq finishes. In this case, mlx5e_alloc_rx_mpwqe may post UMR WQEs to ICOSQ, and with the bad timing, the wqe_info of the first UMR may be overwritten by the wqe_info of the NOP posted by mlx5e_activate_rq.

The consequence is that icosq->db.wqe_info[0].num_wqebbs will be changed from MLX5E_UMR_WQEBBS to 1, disrupting the integrity of the array-based linked list in wqe_info[]. mlx5e_poll_ico_cq will hang in an infinite loop after processing wqe_info[0], because after the corruption, the next item to be processed will be wqe_info[1], which is filled with zeros, and sqcc += wi->num_wqebbs will never move further.

This commit fixes this race condition by using async_icosq to post the NOP and trigger the interrupt. async_icosq is always protected with a spinlock, eliminating the race condition.

Fixes bc77b240b3c5 ("net/mlx5e: Add fragmented memory support for RX multi packet WQE")
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  |  4 ++++
 .../net/ethernet/mellanox/mlx5/core/en/ptp.c  |  1 +
 .../mellanox/mlx5/core/en/reporter_rx.c       |  6 +++++
 .../net/ethernet/mellanox/mlx5/core/en/trap.c |  1 +
 .../ethernet/mellanox/mlx5/core/en/xsk/pool.c |  1 +
 .../mellanox/mlx5/core/en/xsk/setup.c         |  5 +---
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 24 +++++++++++++------
 7 files changed, 31 insertions(+), 11 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en.h +4 -0
index 5ccd6c634274..4c8c8e4c1ef3 100644
@@ -730,6 +730,7 @@ struct mlx5e_rq {
 	u8                     wq_type;
 	u32                    rqn;
 	struct mlx5_core_dev  *mdev;
+	struct mlx5e_channel  *channel;
 	u32  umr_mkey;
 	struct mlx5e_dma_info  wqe_overflow;
 
@@ -1044,6 +1045,9 @@ void mlx5e_close_cq(struct mlx5e_cq *cq);
 int mlx5e_open_locked(struct net_device *netdev);
 int mlx5e_close_locked(struct net_device *netdev);
 
+void mlx5e_trigger_napi_icosq(struct mlx5e_channel *c);
+void mlx5e_trigger_napi_sched(struct napi_struct *napi);
+
 int mlx5e_open_channels(struct mlx5e_priv *priv,
 			struct mlx5e_channels *chs);
 void mlx5e_close_channels(struct mlx5e_channels *chs);
drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c +1 -0
index 82baafd3c00c..fdb82f2b0130 100644
@@ -737,6 +737,7 @@ void mlx5e_ptp_activate_channel(struct mlx5e_ptp *c)
 	if (test_bit(MLX5E_PTP_STATE_RX, c->state)) {
 		mlx5e_ptp_rx_set_fs(c->priv);
 		mlx5e_activate_rq(&c->rq);
+		mlx5e_trigger_napi_sched(&c->napi);
 	}
 }
 
drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c +6 -0
index 2684e9da9f41..fc366e66d0b0 100644
@@ -123,6 +123,8 @@ static int mlx5e_rx_reporter_err_icosq_cqe_recover(void *ctx)
 		xskrq->stats->recover++;
 	}
 
+	mlx5e_trigger_napi_icosq(icosq->channel);
+
 	mutex_unlock(&icosq->channel->icosq_recovery_lock);
 
 	return 0;
@@ -166,6 +168,10 @@ static int mlx5e_rx_reporter_err_rq_cqe_recover(void *ctx)
 	clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state);
 	mlx5e_activate_rq(rq);
 	rq->stats->recover++;
+	if (rq->channel)
+		mlx5e_trigger_napi_icosq(rq->channel);
+	else
+		mlx5e_trigger_napi_sched(rq->cq.napi);
 	return 0;
 out:
 	clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state);
drivers/net/ethernet/mellanox/mlx5/core/en/trap.c +1 -0
index a55b066746cb..6dd36e3cf425 100644
@@ -172,6 +172,7 @@ static void mlx5e_activate_trap(struct mlx5e_trap *trap)
 {
 	napi_enable(&trap->napi);
 	mlx5e_activate_rq(&trap->rq);
+	mlx5e_trigger_napi_sched(&trap->napi);
 }
 
 void mlx5e_deactivate_trap(struct mlx5e_priv *priv)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c +1 -0
index 279cd8f4e79f..2c520394aa1d 100644
@@ -117,6 +117,7 @@ static int mlx5e_xsk_enable_locked(struct mlx5e_priv *priv,
 		goto err_remove_pool;
 
 	mlx5e_activate_xsk(c);
+	mlx5e_trigger_napi_icosq(c);
 
 	/* Don't wait for WQEs, because the newer xdpsock sample doesn't provide
 	 * any Fill Ring entries at the setup stage.
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c +1 -4
index 25eac9e20342..5a2cd15e245d 100644
@@ -64,6 +64,7 @@ static int mlx5e_init_xsk_rq(struct mlx5e_channel *c,
 	rq->clock        = &mdev->clock;
 	rq->icosq        = &c->icosq;
 	rq->ix           = c->ix;
+	rq->channel      = c;
 	rq->mdev         = mdev;
 	rq->hw_mtu       = MLX5E_SW2HW_MTU(params, params->sw_mtu);
 	rq->xdpsq        = &c->rq_xdpsq;
@@ -179,10 +180,6 @@ void mlx5e_activate_xsk(struct mlx5e_channel *c)
 	mlx5e_reporter_icosq_resume_recovery(c);
 
 	/* TX queue is created active. */
-
-	spin_lock_bh(&c->async_icosq_lock);
-	mlx5e_trigger_irq(&c->async_icosq);
-	spin_unlock_bh(&c->async_icosq_lock);
 }
 
 void mlx5e_deactivate_xsk(struct mlx5e_channel *c)
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +17 -7
index 531fffe1abe3..9730bd96d0de 100644
@@ -477,6 +477,7 @@ static int mlx5e_init_rxq_rq(struct mlx5e_channel *c, struct mlx5e_params *param
 	rq->clock        = &mdev->clock;
 	rq->icosq        = &c->icosq;
 	rq->ix           = c->ix;
+	rq->channel      = c;
 	rq->mdev         = mdev;
 	rq->hw_mtu       = MLX5E_SW2HW_MTU(params, params->sw_mtu);
 	rq->xdpsq        = &c->rq_xdpsq;
@@ -1070,13 +1071,6 @@ int mlx5e_open_rq(struct mlx5e_params *params, struct mlx5e_rq_param *param,
 void mlx5e_activate_rq(struct mlx5e_rq *rq)
 {
 	set_bit(MLX5E_RQ_STATE_ENABLED, &rq->state);
-	if (rq->icosq) {
-		mlx5e_trigger_irq(rq->icosq);
-	} else {
-		local_bh_disable();
-		napi_schedule(rq->cq.napi);
-		local_bh_enable();
-	}
 }
 
 void mlx5e_deactivate_rq(struct mlx5e_rq *rq)
@@ -2218,6 +2212,20 @@ static int mlx5e_channel_stats_alloc(struct mlx5e_priv *priv, int ix, int cpu)
 	return 0;
 }
 
+void mlx5e_trigger_napi_icosq(struct mlx5e_channel *c)
+{
+	spin_lock_bh(&c->async_icosq_lock);
+	mlx5e_trigger_irq(&c->async_icosq);
+	spin_unlock_bh(&c->async_icosq_lock);
+}
+
+void mlx5e_trigger_napi_sched(struct napi_struct *napi)
+{
+	local_bh_disable();
+	napi_schedule(napi);
+	local_bh_enable();
+}
+
 static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix,
 			      struct mlx5e_params *params,
 			      struct mlx5e_channel_param *cparam,
@@ -2299,6 +2307,8 @@ static void mlx5e_activate_channel(struct mlx5e_channel *c)
 
 	if (test_bit(MLX5E_CHANNEL_STATE_XSK, c->state))
 		mlx5e_activate_xsk(c);
+
+	mlx5e_trigger_napi_icosq(c);
 }
 
 static void mlx5e_deactivate_channel(struct mlx5e_channel *c)
Signature · 1 line
2.35.1
author maintainer
#112 lore raw

PATCH 112/298 net/mlx5: correct ECE offset in query qp output

In reply to Greg Kroah-Hartman

From: Changcheng Liu <[email protected]>

[ Upstream commit 3fc2a9e89b3508a5cc0c324f26d7b4740ba8c456 ]

ECE field should be after opt_param_mask in query qp output.

Fixes 6b646a7e4af6 ("net/mlx5: Add ability to read and write ECE options")
---
 include/linux/mlx5/mlx5_ifc.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
include/linux/mlx5/mlx5_ifc.h +2 -3
index 49a48d7709ac..4cd54277d5d9 100644
@@ -5175,12 +5175,11 @@ struct mlx5_ifc_query_qp_out_bits {
 
 	u8         syndrome[0x20];
 
-	u8         reserved_at_40[0x20];
-	u8         ece[0x20];
+	u8         reserved_at_40[0x40];
 
 	u8         opt_param_mask[0x20];
 
-	u8         reserved_at_a0[0x20];
+	u8         ece[0x20];
 
 	struct mlx5_ifc_qpc_bits qpc;
 
Signature · 1 line
2.35.1
author maintainer
#113 lore raw

PATCH 113/298 net/mlx5e: Update netdev features after changing XDP state

In reply to Greg Kroah-Hartman

From: Maxim Mikityanskiy <[email protected]>

[ Upstream commit f6279f113ad593971999c877eb69dc3d36a75894 ]

Some features (LRO, HW GRO) conflict with XDP. If there is an attempt to enable such features while XDP is active, they will be set to off [requested on]. In order to activate these features after XDP is turned off, the driver needs to call netdev_update_features(). This commit adds this missing call after XDP state changes.

Fixes cf6e34c8c22f ("net/mlx5e: Properly block LRO when XDP is enabled")
Fixes b0617e7b3500 ("net/mlx5e: Properly block HW GRO when XDP is enabled")
---
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +5 -0
index 9730bd96d0de..352b5c8ae24e 100644
@@ -4531,6 +4531,11 @@ static int mlx5e_xdp_set(struct net_device *netdev, struct bpf_prog *prog)
 
 unlock:
 	mutex_unlock(&priv->state_lock);
+
+	/* Need to fix some features. */
+	if (!err)
+		netdev_update_features(netdev);
+
 	return err;
 }
 
Signature · 1 line
2.35.1
author maintainer
#114 lore raw

PATCH 114/298 net: sched: add barrier to fix packet stuck problem for lockless qdisc

In reply to Greg Kroah-Hartman

From: Guoju Fang <[email protected]>

[ Upstream commit 2e8728c955ce0624b958eee6e030a37aca3a5d86 ]

In qdisc_run_end(), the spin_unlock() only has store-release semantic, which guarantees all earlier memory access are visible before it. But the subsequent test_bit() has no barrier semantics so may be reordered ahead of the spin_unlock(). The store-load reordering may cause a packet stuck problem.

The concurrent operations can be described as below,

         CPU 0                      |          CPU 1
   qdisc_run_end()                  |     qdisc_run_begin()
          .                         |           .
 ----> /* may be reorderd here */   |           .
|         .                         |           .
|     spin_unlock()                 |         set_bit()
|         .                         |         smp_mb__after_atomic()
 ---- test_bit()                    |         spin_trylock()
          .                         |          .

Consider the following sequence of events:
CPU 0 reorder test_bit() ahead and see MISSED = 0 CPU 1 calls set_bit()
CPU 1 calls spin_trylock() and return fail CPU 0 executes spin_unlock()

At the end of the sequence, CPU 0 calls spin_unlock() and does nothing because it see MISSED = 0. The skb on CPU 1 has beed enqueued but no one take it, until the next cpu pushing to the qdisc (if ever ...) will notice and dequeue it.

This patch fix this by adding one explicit barrier. As spin_unlock() and test_bit() ordering is a store-load ordering, a full memory barrier smp_mb() is needed here.

Fixes a90c57f2cedd ("net: sched: fix packet stuck problem for lockless qdisc")
---
 include/net/sch_generic.h | 6 ++++++
 1 file changed, 6 insertions(+)
include/net/sch_generic.h +6 -0
index 4fe0892dde1a..6764fc265745 100644
@@ -209,6 +209,12 @@ static inline void qdisc_run_end(struct Qdisc *qdisc)
 	if (qdisc->flags & TCQ_F_NOLOCK) {
 		spin_unlock(&qdisc->seqlock);
 
+		/* spin_unlock() only has store-release semantic. The unlock
+		 * and test_bit() ordering is a store-load ordering, so a full
+		 * memory barrier is needed here.
+		 */
+		smp_mb();
+
 		if (unlikely(test_bit(__QDISC_STATE_MISSED,
 				      &qdisc->state)))
 			__netif_schedule(qdisc);
Signature · 1 line
2.35.1
author maintainer
#115 lore raw

PATCH 115/298 tcp: tcp_rtx_synack() can be called from process context

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit 0a375c822497ed6ad6b5da0792a12a6f1af10c0b ]

Laurent reported the enclosed report [1]

This bug triggers with following coditions:

0) Kernel built with CONFIG_DEBUG_PREEMPT=y

1) A new passive FastOpen TCP socket is created.
This FO socket waits for an ACK coming from client to be a complete ESTABLISHED one.
2) A socket operation on this socket goes through lock_sock() release_sock() dance.
3) While the socket is owned by the user in step 2),
a retransmit of the SYN is received and stored in socket backlog.
4) At release_sock() time, the socket backlog is processed while in process context.
5) A SYNACK packet is cooked in response of the SYN retransmit.
6) -> tcp_rtx_synack() is called in process context.

Before blamed commit, tcp_rtx_synack() was always called from BH handler, from a timer handler.

Fix this by using TCP_INC_STATS() & NET_INC_STATS() which do not assume caller is in non preemptible context.

[1]

BUG: using __this_cpu_add() in preemptible [00000000] code: epollpep/2180
caller is tcp_rtx_synack.part.0+0x36/0xc0
CPU: 10 PID: 2180 Comm: epollpep Tainted: G           OE     5.16.0-0.bpo.4-amd64 #1  Debian 5.16.12-1~bpo11+1
Hardware name: Supermicro SYS-5039MC-H8TRF/X11SCD-F, BIOS 1.7 11/23/2021
Call Trace:
 <TASK>
 dump_stack_lvl+0x48/0x5e
 check_preemption_disabled+0xde/0xe0
 tcp_rtx_synack.part.0+0x36/0xc0
 tcp_rtx_synack+0x8d/0xa0
 ? kmem_cache_alloc+0x2e0/0x3e0
 ? apparmor_file_alloc_security+0x3b/0x1f0
 inet_rtx_syn_ack+0x16/0x30
 tcp_check_req+0x367/0x610
 tcp_rcv_state_process+0x91/0xf60
 ? get_nohz_timer_target+0x18/0x1a0
 ? lock_timer_base+0x61/0x80
 ? preempt_count_add+0x68/0xa0
 tcp_v4_do_rcv+0xbd/0x270
 __release_sock+0x6d/0xb0
 release_sock+0x2b/0x90
 sock_setsockopt+0x138/0x1140
 ? __sys_getsockname+0x7e/0xc0
 ? aa_sk_perm+0x3e/0x1a0
 __sys_setsockopt+0x198/0x1e0
 __x64_sys_setsockopt+0x21/0x30
 do_syscall_64+0x38/0xc0
 entry_SYSCALL_64_after_hwframe+0x44/0xae
---
 net/ipv4/tcp_output.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/ipv4/tcp_output.c +2 -2
index 84651fc3dde0..2adff4877cd6 100644
@@ -4100,8 +4100,8 @@ int tcp_rtx_synack(const struct sock *sk, struct request_sock *req)
 	res = af_ops->send_synack(sk, NULL, &fl, req, NULL, TCP_SYNACK_NORMAL,
 				  NULL);
 	if (!res) {
-		__TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS);
-		__NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS);
+		TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS);
+		NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS);
 		if (unlikely(tcp_passive_fastopen(sk)))
 			tcp_sk(sk)->total_retrans++;
 		trace_tcp_retransmit_synack(sk, req);
Signature · 1 line
2.35.1
author maintainer
#116 lore raw

PATCH 116/298 vdpa: ifcvf: set pci driver data in probe

In reply to Greg Kroah-Hartman

From: Jason Wang <[email protected]>

[ Upstream commit bd8bb9aed56b1814784a975e2dfea12a9adcee92 ]

We should set the pci driver data in probe instead of the vdpa device adding callback. Otherwise if no vDPA device is created we will lose the pointer to the management device.

Fixes 6b5df347c6482 ("vDPA/ifcvf: implement management netlink framework for ifcvf")

Message-Id: <[email protected]>

---
 drivers/vdpa/ifcvf/ifcvf_main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/vdpa/ifcvf/ifcvf_main.c +2 -1
index d1a6b5ab543c..474c6120c955 100644
@@ -514,7 +514,6 @@ static int ifcvf_vdpa_dev_add(struct vdpa_mgmt_dev *mdev, const char *name,
 	}
 
 	ifcvf_mgmt_dev->adapter = adapter;
-	pci_set_drvdata(pdev, ifcvf_mgmt_dev);
 
 	vf = &adapter->vf;
 	vf->dev_type = get_dev_type(pdev);
@@ -629,6 +628,8 @@ static int ifcvf_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		goto err;
 	}
 
+	pci_set_drvdata(pdev, ifcvf_mgmt_dev);
+
 	return 0;
 
 err:
Signature · 1 line
2.35.1
author maintainer
#117 lore raw

PATCH 117/298 octeontx2-af: fix error code in is_valid_offset()

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit f3d671c711097a133bc36bd2bde52f1fcca783a6 ]

The is_valid_offset() function returns success/true if the call to validate_and_get_cpt_blkaddr() fails.

Fixes ecad2ce8c48f ("octeontx2-af: cn10k: Add mailbox to configure reassembly timeout")
---
 drivers/net/ethernet/marvell/octeontx2/af/rvu_cpt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/marvell/octeontx2/af/rvu_cpt.c +1 -1
index a73a8017e0ee..e3a317442c8c 100644
@@ -579,7 +579,7 @@ static bool is_valid_offset(struct rvu *rvu, struct cpt_rd_wr_reg_msg *req)
 
 	blkaddr = validate_and_get_cpt_blkaddr(req->blkaddr);
 	if (blkaddr < 0)
-		return blkaddr;
+		return false;
 
 	/* Registers that can be accessed from PF/VF */
 	if ((offset & 0xFF000) ==  CPT_AF_LFX_CTL(0) ||
Signature · 1 line
2.35.1
author maintainer
#118 lore raw

PATCH 118/298 macsec: fix UAF bug for real_dev

In reply to Greg Kroah-Hartman

From: Ziyang Xuan <[email protected]>

[ Upstream commit 196a888ca6571deb344468e1d7138e3273206335 ]

Create a new macsec device but not get reference to real_dev. That can not ensure that real_dev is freed after macsec. That will trigger the UAF bug for real_dev as following:

==================================================================

BUG: KASAN: use-after-free in macsec_get_iflink+0x5f/0x70 drivers/net/macsec.c:3662
Call Trace:
 ...
 macsec_get_iflink+0x5f/0x70 drivers/net/macsec.c:3662
 dev_get_iflink+0x73/0xe0 net/core/dev.c:637
 default_operstate net/core/link_watch.c:42 [inline]
 rfc2863_policy+0x233/0x2d0 net/core/link_watch.c:54
 linkwatch_do_dev+0x2a/0x150 net/core/link_watch.c:161

Allocated by task 22209:
 ...
 alloc_netdev_mqs+0x98/0x1100 net/core/dev.c:10549
 rtnl_create_link+0x9d7/0xc00 net/core/rtnetlink.c:3235
 veth_newlink+0x20e/0xa90 drivers/net/veth.c:1748

Freed by task 8:
 ...
 kfree+0xd6/0x4d0 mm/slub.c:4552
 kvfree+0x42/0x50 mm/util.c:615
 device_release+0x9f/0x240 drivers/base/core.c:2229
 kobject_cleanup lib/kobject.c:673 [inline]
 kobject_release lib/kobject.c:704 [inline]
 kref_put include/linux/kref.h:65 [inline]
 kobject_put+0x1c8/0x540 lib/kobject.c:721
 netdev_run_todo+0x72e/0x10b0 net/core/dev.c:10327

After commit faab39f63c1f ("net: allow out-of-order netdev unregistration") and commit e5f80fcf869a ("ipv6: give an IPv6 dev to blackhole_netdev"), we can add dev_hold_track() in macsec_dev_init() and dev_put_track() in macsec_free_netdev() to fix the problem.

Fixes 2bce1ebed17d ("macsec: fix refcnt leak in module exit routine")
Reported-by [email protected]
---
 drivers/net/macsec.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/net/macsec.c +7 -0
index 3d0874331763..b901acca098b 100644
@@ -99,6 +99,7 @@ struct pcpu_secy_stats {
  * struct macsec_dev - private data
  * @secy: SecY config
  * @real_dev: pointer to underlying netdevice
+ * @dev_tracker: refcount tracker for @real_dev reference
  * @stats: MACsec device stats
  * @secys: linked list of SecY's on the underlying device
  * @gro_cells: pointer to the Generic Receive Offload cell
@@ -107,6 +108,7 @@ struct pcpu_secy_stats {
 struct macsec_dev {
 	struct macsec_secy secy;
 	struct net_device *real_dev;
+	netdevice_tracker dev_tracker;
 	struct pcpu_secy_stats __percpu *stats;
 	struct list_head secys;
 	struct gro_cells gro_cells;
@@ -3459,6 +3461,9 @@ static int macsec_dev_init(struct net_device *dev)
 	if (is_zero_ether_addr(dev->broadcast))
 		memcpy(dev->broadcast, real_dev->broadcast, dev->addr_len);
 
+	/* Get macsec's reference to real_dev */
+	dev_hold_track(real_dev, &macsec->dev_tracker, GFP_KERNEL);
+
 	return 0;
 }
 
@@ -3704,6 +3709,8 @@ static void macsec_free_netdev(struct net_device *dev)
 	free_percpu(macsec->stats);
 	free_percpu(macsec->secy.tx_sc.stats);
 
+	/* Get rid of the macsec's reference to real_dev */
+	dev_put_track(macsec->real_dev, &macsec->dev_tracker);
 }
 
 static void macsec_setup(struct net_device *dev)
Signature · 1 line
2.35.1
author maintainer
#119 lore raw

PATCH 119/298 s390/mcck: isolate SIE instruction when setting CIF_MCCK_GUEST flag

In reply to Greg Kroah-Hartman

From: Alexander Gordeev <[email protected]>

[ Upstream commit 29ccaa4b35ea874ddd50518e5c2c746b9238a792 ]

Commit d768bd892fc8 ("s390: add options to change branch prediction behaviour for the kernel") introduced .Lsie_exit label - supposedly to fence off SIE instruction. However, the corresponding address range length .Lsie_crit_mcck_length was not updated, which led to BPON code potentionally marked with CIF_MCCK_GUEST flag.

Both .Lsie_exit and .Lsie_crit_mcck_length were removed with commit 0b0ed657fe00 ("s390: remove critical section cleanup from entry.S"), but the issue persisted - currently BPOFF and BPENTER macros might get wrongly considered by the machine check handler as a guest.

Fixes d768bd892fc8 ("s390: add options to change branch prediction behaviour for the kernel")
---
 arch/s390/kernel/entry.S | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
arch/s390/kernel/entry.S +5 -1
index 01bae1d51113..3bf8aeeec96f 100644
@@ -264,6 +264,10 @@ ENTRY(sie64a)
 	BPEXIT	__SF_SIE_FLAGS(%r15),(_TIF_ISOLATE_BP|_TIF_ISOLATE_BP_GUEST)
 .Lsie_entry:
 	sie	0(%r14)
+# Let the next instruction be NOP to avoid triggering a machine check
+# and handling it in a guest as result of the instruction execution.
+	nopr	7
+.Lsie_leave:
 	BPOFF
 	BPENTER	__SF_SIE_FLAGS(%r15),(_TIF_ISOLATE_BP|_TIF_ISOLATE_BP_GUEST)
 .Lsie_skip:
@@ -563,7 +567,7 @@ ENTRY(mcck_int_handler)
 	jno	.Lmcck_panic
 #if IS_ENABLED(CONFIG_KVM)
 	OUTSIDE	%r9,.Lsie_gmap,.Lsie_done,6f
-	OUTSIDE	%r9,.Lsie_entry,.Lsie_skip,4f
+	OUTSIDE	%r9,.Lsie_entry,.Lsie_leave,4f
 	oi	__LC_CPU_FLAGS+7, _CIF_MCCK_GUEST
 	j	5f
 4:	CHKSTG	.Lmcck_panic
Signature · 1 line
2.35.1
author maintainer
#120 lore raw

PATCH 120/298 regulator: mt6315-regulator: fix invalid allowed mode

In reply to Greg Kroah-Hartman

From: Fabien Parent <[email protected]>

[ Upstream commit 28cbc2d4c54c09a427b18a1604740efb6b2cc2d6 ]

In the binding example, the regulator mode 4 is shown as a valid mode, but the driver actually only support mode 0 to 2:

This generates an error in dmesg when copy/pasting the binding example:

[    0.306080] vbuck1: invalid regulator-allowed-modes element 4
[    0.307290] vbuck2: invalid regulator-allowed-modes element 4

This commit fixes this error by removing the invalid mode from the examples.

Fixes 977fb5b58469 ("regulator: document binding for MT6315 regulator")
---
 .../devicetree/bindings/regulator/mt6315-regulator.yaml       | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
Documentation/devicetree/bindings/regulator/mt6315-regulator.yaml +2 -2
index 5d2d989de893..37402c370fbb 100644
@@ -55,7 +55,7 @@ examples:
           regulator-min-microvolt = <300000>;
           regulator-max-microvolt = <1193750>;
           regulator-enable-ramp-delay = <256>;
-          regulator-allowed-modes = <0 1 2 4>;
+          regulator-allowed-modes = <0 1 2>;
         };
 
         vbuck3 {
@@ -63,7 +63,7 @@ examples:
           regulator-min-microvolt = <300000>;
           regulator-max-microvolt = <1193750>;
           regulator-enable-ramp-delay = <256>;
-          regulator-allowed-modes = <0 1 2 4>;
+          regulator-allowed-modes = <0 1 2>;
         };
       };
     };
Signature · 1 line
2.35.1
author maintainer
#121 lore raw

PATCH 121/298 gpio: pca953x: use the correct register address to do regcache sync

In reply to Greg Kroah-Hartman

From: Haibo Chen <[email protected]>

[ Upstream commit 43624eda86c98b0de726d0b6f2516ccc3ef7313f ]

For regcache_sync_region, need to use pca953x_recalc_addr() to get the real register address.

Fixes b76574300504 ("gpio: pca953x: Restore registers after suspend/resume cycle")
---
 drivers/gpio/gpio-pca953x.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)
drivers/gpio/gpio-pca953x.c +11 -8
index 8726921a1129..33683295a0bf 100644
@@ -1108,20 +1108,21 @@ static int pca953x_regcache_sync(struct device *dev)
 {
 	struct pca953x_chip *chip = dev_get_drvdata(dev);
 	int ret;
+	u8 regaddr;
 
 	/*
 	 * The ordering between direction and output is important,
 	 * sync these registers first and only then sync the rest.
 	 */
-	ret = regcache_sync_region(chip->regmap, chip->regs->direction,
-				   chip->regs->direction + NBANK(chip));
+	regaddr = pca953x_recalc_addr(chip, chip->regs->direction, 0);
+	ret = regcache_sync_region(chip->regmap, regaddr, regaddr + NBANK(chip));
 	if (ret) {
 		dev_err(dev, "Failed to sync GPIO dir registers: %d\n", ret);
 		return ret;
 	}
 
-	ret = regcache_sync_region(chip->regmap, chip->regs->output,
-				   chip->regs->output + NBANK(chip));
+	regaddr = pca953x_recalc_addr(chip, chip->regs->output, 0);
+	ret = regcache_sync_region(chip->regmap, regaddr, regaddr + NBANK(chip));
 	if (ret) {
 		dev_err(dev, "Failed to sync GPIO out registers: %d\n", ret);
 		return ret;
@@ -1129,16 +1130,18 @@ static int pca953x_regcache_sync(struct device *dev)
 
 #ifdef CONFIG_GPIO_PCA953X_IRQ
 	if (chip->driver_data & PCA_PCAL) {
-		ret = regcache_sync_region(chip->regmap, PCAL953X_IN_LATCH,
-					   PCAL953X_IN_LATCH + NBANK(chip));
+		regaddr = pca953x_recalc_addr(chip, PCAL953X_IN_LATCH, 0);
+		ret = regcache_sync_region(chip->regmap, regaddr,
+					   regaddr + NBANK(chip));
 		if (ret) {
 			dev_err(dev, "Failed to sync INT latch registers: %d\n",
 				ret);
 			return ret;
 		}
 
-		ret = regcache_sync_region(chip->regmap, PCAL953X_INT_MASK,
-					   PCAL953X_INT_MASK + NBANK(chip));
+		regaddr = pca953x_recalc_addr(chip, PCAL953X_INT_MASK, 0);
+		ret = regcache_sync_region(chip->regmap, regaddr,
+					   regaddr + NBANK(chip));
 		if (ret) {
 			dev_err(dev, "Failed to sync INT mask registers: %d\n",
 				ret);
Signature · 1 line
2.35.1
author maintainer
#122 lore raw

PATCH 122/298 afs: Fix infinite loop found by xfstest generic/676

In reply to Greg Kroah-Hartman

From: David Howells <[email protected]>

[ Upstream commit 17eabd42560f4636648ad65ba5b20228071e2363 ]

In AFS, a directory is handled as a file that the client downloads and parses locally for the purposes of performing lookup and getdents operations. The in-kernel afs filesystem has a number of functions that do this.

A directory file is arranged as a series of 2K blocks divided into 32-byte slots, where a directory entry occupies one or more slots, plus each block starts with one or more metadata blocks.

When parsing a block, if the last slots are occupied by a dirent that occupies more than a single slot and the file position points at a slot that's not the initial one, the logic in afs_dir_iterate_block() that skips over it won't advance the file pointer to the end of it. This will cause an infinite loop in getdents() as it will keep retrying that block and failing to advance beyond the final entry.

Fix this by advancing the file pointer if the next entry will be beyond it when we skip a block.

This was found by the generic/676 xfstest but can also be triggered with something like:

	~/xfstests-dev/src/t_readdir_3 /xfstest.test/z 4000 1
---
 fs/afs/dir.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
fs/afs/dir.c +4 -1
index da9b4f8577a1..7f1cb3b73874 100644
@@ -462,8 +462,11 @@ static int afs_dir_iterate_block(struct afs_vnode *dvnode,
 		}
 
 		/* skip if starts before the current position */
-		if (offset < curr)
+		if (offset < curr) {
+			if (next > curr)
+				ctx->pos = blkoff + next * sizeof(union afs_xdr_dirent);
 			continue;
+		}
 
 		/* found the next entry */
 		if (!dir_emit(ctx, dire->u.name, nlen,
Signature · 1 line
2.35.1
author maintainer
#123 lore raw

PATCH 123/298 drm/msm/dp: Always clear mask bits to disable interrupts at dp_ctrl_reset_irq_ctrl()

In reply to Greg Kroah-Hartman

From: Kuogee Hsieh <[email protected]>

[ Upstream commit 993a2adc6e2e94a0a7b5bfc054eda90ac95f62c3 ]

dp_catalog_ctrl_reset() will software reset DP controller. But it will not reset programmable registers to default value. DP driver still have to clear mask bits to interrupt status registers to disable interrupts after software reset of controller.

At current implementation, dp_ctrl_reset_irq_ctrl() will software reset dp controller but did not call dp_catalog_ctrl_enable_irq(false) to clear hpd related interrupt mask bits to disable hpd related interrupts due to it mistakenly think hpd related interrupt mask bits will be cleared by software reset of dp controller automatically. This mistake may cause system to crash during suspending procedure due to unexpected irq fired and trigger event thread to access dp controller registers with controller clocks are disabled.

This patch fixes system crash during suspending problem by removing "enable" flag condition checking at dp_ctrl_reset_irq_ctrl() so that hpd related interrupt mask bits are cleared to prevent unexpected from happening.

Changes in v2:
-- add more details commit text

Changes in v3:
-- add synchrons_irq()
-- add atomic_t suspended

Changes in v4:
-- correct Fixes's commit ID
-- remove synchrons_irq()

Changes in v5:
-- revise commit text

Changes in v6:
-- add event_lock to protect "suspended"

Changes in v7:
-- delete "suspended" flag

Fixes 989ebe7bc446 ("drm/msm/dp: do not initialize phy until plugin interrupt received")

Patchwork: https://patchwork.freedesktop.org/patch/486591/

---
 drivers/gpu/drm/msm/dp/dp_ctrl.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/gpu/drm/msm/dp/dp_ctrl.c +7 -2
index 6eb176872a17..7ae74bd05924 100644
@@ -1373,8 +1373,13 @@ void dp_ctrl_reset_irq_ctrl(struct dp_ctrl *dp_ctrl, bool enable)
 
 	dp_catalog_ctrl_reset(ctrl->catalog);
 
-	if (enable)
-		dp_catalog_ctrl_enable_irq(ctrl->catalog, enable);
+	/*
+	 * all dp controller programmable registers will not
+	 * be reset to default value after DP_SW_RESET
+	 * therefore interrupt mask bits have to be updated
+	 * to enable/disable interrupts
+	 */
+	dp_catalog_ctrl_enable_irq(ctrl->catalog, enable);
 }
 
 void dp_ctrl_phy_init(struct dp_ctrl *dp_ctrl)
Signature · 1 line
2.35.1
author maintainer
#124 lore raw

PATCH 124/298 scsi: sd: Fix potential NULL pointer dereference

In reply to Greg Kroah-Hartman

From: Damien Le Moal <[email protected]>

[ Upstream commit 05fbde3a77a4f1d62e4c4428f384288c1f1a0be5 ]

If sd_probe() sees an early error before sdkp->device is initialized, sd_zbc_release_disk() is called. This causes a NULL pointer dereference when sd_is_zoned() is called inside that function. Avoid this by removing the call to sd_zbc_release_disk() in sd_probe() error path.

This change is safe and does not result in zone information memory leakage because the zone information for a zoned disk is allocated only when sd_revalidate_disk() is called, at which point sdkp->disk_dev is fully set, resulting in sd_disk_release() being called when needed to cleanup a disk zone information using sd_zbc_release_disk().

---
 drivers/scsi/sd.c | 1 -
 1 file changed, 1 deletion(-)
drivers/scsi/sd.c +0 -1
index 8b5d2a4076c2..65b501859141 100644
@@ -3605,7 +3605,6 @@ static int sd_probe(struct device *dev)
  out_put:
 	put_disk(gd);
  out_free:
-	sd_zbc_release_disk(sdkp);
 	kfree(sdkp);
  out:
 	scsi_autopm_put_device(sdp);
Signature · 1 line
2.35.1
author maintainer
#125 lore raw

PATCH 125/298 ax25: Fix ax25 session cleanup problems

In reply to Greg Kroah-Hartman

From: Duoming Zhou <[email protected]>

[ Upstream commit 7d8a3a477b3e25ada8dc71d22048c2ea417209a0 ]

There are session cleanup problems in ax25_release() and ax25_disconnect(). If we setup a session and then disconnect, the disconnected session is still in "LISTENING" state that is shown below.

Active AX.25 sockets

Dest       Source     Device  State        Vr/Vs    Send-Q  Recv-Q
DL9SAU-4   DL9SAU-3   ???     LISTENING    000/000  0       0
DL9SAU-3   DL9SAU-4   ???     LISTENING    000/000  0       0

The first reason is caused by del_timer_sync() in ax25_release(). The timers of ax25 are used for correct session cleanup. If we use ax25_release() to close ax25 sessions and ax25_dev is not null, the del_timer_sync() functions in ax25_release() will execute. As a result, the sessions could not be cleaned up correctly, because the timers have stopped.

In order to solve this problem, this patch adds a device_up flag in ax25_dev in order to judge whether the device is up. If there are sessions to be cleaned up, the del_timer_sync() in ax25_release() will not execute. What's more, we add ax25_cb_del() in ax25_kill_by_device(), because the timers have been stopped and there are no functions that could delete ax25_cb if we do not call ax25_release(). Finally, we reorder the position of ax25_list_lock in ax25_cb_del() in order to synchronize among different functions that call ax25_cb_del().

The second reason is caused by improper check in ax25_disconnect(). The incoming ax25 sessions which ax25->sk is null will close heartbeat timer, because the check "if(!ax25->sk || ..)" is satisfied. As a result, the session could not be cleaned up properly.

In order to solve this problem, this patch changes the improper check to "if(ax25->sk && ..)" in ax25_disconnect().

What`s more, the ax25_disconnect() may be called twice, which is not necessary. For example, ax25_kill_by_device() calls ax25_disconnect() and sets ax25->state to AX25_STATE_0, but ax25_release() calls ax25_disconnect() again.

In order to solve this problem, this patch add a check in ax25_release(). If the flag of ax25->sk equals to SOCK_DEAD, the ax25_disconnect() in ax25_release() should not be executed.

Fixes 82e31755e55f ("ax25: Fix UAF bugs in ax25 timers")
Fixes 8a367e74c012 ("ax25: Fix segfault after sock connection timeout")
---
 include/net/ax25.h   |  1 +
 net/ax25/af_ax25.c   | 27 +++++++++++++++++----------
 net/ax25/ax25_dev.c  |  1 +
 net/ax25/ax25_subr.c |  2 +-
 4 files changed, 20 insertions(+), 11 deletions(-)
include/net/ax25.h +1 -0
index 8221af1811df..5253692db9eb 100644
@@ -240,6 +240,7 @@ typedef struct ax25_dev {
 	ax25_dama_info		dama;
 #endif
 	refcount_t		refcount;
+	bool device_up;
 } ax25_dev;
 
 typedef struct ax25_cb {
net/ax25/af_ax25.c +17 -10
index 363d47f94532..289f355e1853 100644
@@ -62,12 +62,12 @@ static void ax25_free_sock(struct sock *sk)
  */
 static void ax25_cb_del(ax25_cb *ax25)
 {
+	spin_lock_bh(&ax25_list_lock);
 	if (!hlist_unhashed(&ax25->ax25_node)) {
-		spin_lock_bh(&ax25_list_lock);
 		hlist_del_init(&ax25->ax25_node);
-		spin_unlock_bh(&ax25_list_lock);
 		ax25_cb_put(ax25);
 	}
+	spin_unlock_bh(&ax25_list_lock);
 }
 
 /*
@@ -81,6 +81,7 @@ static void ax25_kill_by_device(struct net_device *dev)
 
 	if ((ax25_dev = ax25_dev_ax25dev(dev)) == NULL)
 		return;
+	ax25_dev->device_up = false;
 
 	spin_lock_bh(&ax25_list_lock);
 again:
@@ -91,6 +92,7 @@ static void ax25_kill_by_device(struct net_device *dev)
 				spin_unlock_bh(&ax25_list_lock);
 				ax25_disconnect(s, ENETUNREACH);
 				s->ax25_dev = NULL;
+				ax25_cb_del(s);
 				spin_lock_bh(&ax25_list_lock);
 				goto again;
 			}
@@ -103,6 +105,7 @@ static void ax25_kill_by_device(struct net_device *dev)
 				dev_put_track(ax25_dev->dev, &ax25_dev->dev_tracker);
 				ax25_dev_put(ax25_dev);
 			}
+			ax25_cb_del(s);
 			release_sock(sk);
 			spin_lock_bh(&ax25_list_lock);
 			sock_put(sk);
@@ -995,9 +998,11 @@ static int ax25_release(struct socket *sock)
 	if (sk->sk_type == SOCK_SEQPACKET) {
 		switch (ax25->state) {
 		case AX25_STATE_0:
-			release_sock(sk);
-			ax25_disconnect(ax25, 0);
-			lock_sock(sk);
+			if (!sock_flag(ax25->sk, SOCK_DEAD)) {
+				release_sock(sk);
+				ax25_disconnect(ax25, 0);
+				lock_sock(sk);
+			}
 			ax25_destroy_socket(ax25);
 			break;
 
@@ -1053,11 +1058,13 @@ static int ax25_release(struct socket *sock)
 		ax25_destroy_socket(ax25);
 	}
 	if (ax25_dev) {
-		del_timer_sync(&ax25->timer);
-		del_timer_sync(&ax25->t1timer);
-		del_timer_sync(&ax25->t2timer);
-		del_timer_sync(&ax25->t3timer);
-		del_timer_sync(&ax25->idletimer);
+		if (!ax25_dev->device_up) {
+			del_timer_sync(&ax25->timer);
+			del_timer_sync(&ax25->t1timer);
+			del_timer_sync(&ax25->t2timer);
+			del_timer_sync(&ax25->t3timer);
+			del_timer_sync(&ax25->idletimer);
+		}
 		dev_put_track(ax25_dev->dev, &ax25_dev->dev_tracker);
 		ax25_dev_put(ax25_dev);
 	}
net/ax25/ax25_dev.c +1 -0
index d2a244e1c260..5451be15e072 100644
@@ -62,6 +62,7 @@ void ax25_dev_device_up(struct net_device *dev)
 	ax25_dev->dev     = dev;
 	dev_hold_track(dev, &ax25_dev->dev_tracker, GFP_ATOMIC);
 	ax25_dev->forward = NULL;
+	ax25_dev->device_up = true;
 
 	ax25_dev->values[AX25_VALUES_IPDEFMODE] = AX25_DEF_IPDEFMODE;
 	ax25_dev->values[AX25_VALUES_AXDEFMODE] = AX25_DEF_AXDEFMODE;
net/ax25/ax25_subr.c +1 -1
index 3a476e4f6cd0..9ff98f46dc6b 100644
@@ -268,7 +268,7 @@ void ax25_disconnect(ax25_cb *ax25, int reason)
 		del_timer_sync(&ax25->t3timer);
 		del_timer_sync(&ax25->idletimer);
 	} else {
-		if (!ax25->sk || !sock_flag(ax25->sk, SOCK_DESTROY))
+		if (ax25->sk && !sock_flag(ax25->sk, SOCK_DESTROY))
 			ax25_stop_heartbeat(ax25);
 		ax25_stop_t1timer(ax25);
 		ax25_stop_t2timer(ax25);
Signature · 1 line
2.35.1
author maintainer
#126 lore raw

PATCH 126/298 tipc: check attribute length for bearer name

In reply to Greg Kroah-Hartman

From: Hoang Le <[email protected]>

[ Upstream commit 7f36f798f89bf32c0164049cb0e3fd1af613d0bb ]

syzbot reported uninit-value:
=====================================================

BUG: KMSAN: uninit-value in string_nocheck lib/vsprintf.c:644 [inline]
BUG: KMSAN: uninit-value in string+0x4f9/0x6f0 lib/vsprintf.c:725
 string_nocheck lib/vsprintf.c:644 [inline]
 string+0x4f9/0x6f0 lib/vsprintf.c:725
 vsnprintf+0x2222/0x3650 lib/vsprintf.c:2806
 vprintk_store+0x537/0x2150 kernel/printk/printk.c:2158
 vprintk_emit+0x28b/0xab0 kernel/printk/printk.c:2256
 vprintk_default+0x86/0xa0 kernel/printk/printk.c:2283
 vprintk+0x15f/0x180 kernel/printk/printk_safe.c:50
 _printk+0x18d/0x1cf kernel/printk/printk.c:2293
 tipc_enable_bearer net/tipc/bearer.c:371 [inline]
 __tipc_nl_bearer_enable+0x2022/0x22a0 net/tipc/bearer.c:1033
 tipc_nl_bearer_enable+0x6c/0xb0 net/tipc/bearer.c:1042
 genl_family_rcv_msg_doit net/netlink/genetlink.c:731 [inline]

- Do sanity check the attribute length for TIPC_NLA_BEARER_NAME.
- Do not use 'illegal name' in printing message.

Reported-by [email protected]
Fixes cb30a63384bc ("tipc: refactor function tipc_enable_bearer()")
---
 net/tipc/bearer.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
net/tipc/bearer.c +1 -2
index a2f9c9640716..91d9c815a406 100644
@@ -259,9 +259,8 @@ static int tipc_enable_bearer(struct net *net, const char *name,
 	u32 i;
 
 	if (!bearer_name_validate(name, &b_names)) {
-		errstr = "illegal name";
 		NL_SET_ERR_MSG(extack, "Illegal name");
-		goto rejected;
+		return res;
 	}
 
 	if (prio > TIPC_MAX_LINK_PRI && prio != TIPC_MEDIA_LINK_PRI) {
Signature · 1 line
2.35.1
author maintainer
#127 lore raw

PATCH 127/298 driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

In reply to Greg Kroah-Hartman

From: Saravana Kannan <[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] - https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] - https://lore.kernel.org/lkml/[email protected]/
[3] - https://lore.kernel.org/lkml/[email protected]/

Fixes 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires")
Cc John Stultz <[email protected]>
Cc "David S. Miller" <[email protected]>
Cc Alexey Kuznetsov <[email protected]>
Cc Hideaki YOSHIFUJI <[email protected]>
Cc Jakub Kicinski <[email protected]>
Cc Rob Herring <[email protected]>
Cc Geert Uytterhoeven <[email protected]>
Cc Yoshihiro Shimoda <[email protected]>
Cc Robin Murphy <[email protected]>
Cc Andy Shevchenko <[email protected]>
Cc Sudeep Holla <[email protected]>
Cc Andy Shevchenko <[email protected]>
Cc Naresh Kamboju <[email protected]>
Cc Basil Eljuse <[email protected]>
Cc Ferry Toth <[email protected]>
Cc Arnd Bergmann <[email protected]>
Cc Anders Roxell <[email protected]>
---
 drivers/base/dd.c | 5 -----
 1 file changed, 5 deletions(-)
drivers/base/dd.c +0 -5
index 977e94cf669e..86fd2ea35656 100644
@@ -257,7 +257,6 @@ DEFINE_SHOW_ATTRIBUTE(deferred_devs);
 
 int driver_deferred_probe_timeout;
 EXPORT_SYMBOL_GPL(driver_deferred_probe_timeout);
-static DECLARE_WAIT_QUEUE_HEAD(probe_timeout_waitqueue);
 
 static int __init deferred_probe_timeout_setup(char *str)
 {
@@ -312,7 +311,6 @@ static void deferred_probe_timeout_work_func(struct work_struct *work)
 	list_for_each_entry(p, &deferred_probe_pending_list, deferred_probe)
 		dev_info(p->device, "deferred probe pending\n");
 	mutex_unlock(&deferred_probe_mutex);
-	wake_up_all(&probe_timeout_waitqueue);
 }
 static DECLARE_DELAYED_WORK(deferred_probe_timeout_work, deferred_probe_timeout_work_func);
 
@@ -720,9 +718,6 @@ int driver_probe_done(void)
  */
 void wait_for_device_probe(void)
 {
-	/* wait for probe timeout */
-	wait_event(probe_timeout_waitqueue, !driver_deferred_probe_timeout);
-
 	/* wait for the deferred probe workqueue to finish */
 	flush_work(&deferred_probe_work);
 
Signature · 1 line
2.35.1
Tested-by Geert Uytterhoeven Acked-by John Stultz Reviewed-by Rafael J. Wysocki on v1 127/298
#128 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

Summary:
Start: 686c84f2f136 Linux 5.10.189-rc1
Plain log: https://storage.kernelci.org/stable-rc/linux-5.10.y/v5.10.188-183-g686c84f2f1364/arm64/defconfig/gcc-10/lab-collabora/baseline-r8a77960-ulcb.txt
HTML log: https://storage.kernelci.org/stable-rc/linux-5.10.y/v5.10.188-183-g686c84f2f1364/arm64/defconfig/gcc-10/lab-collabora/baseline-r8a77960-ulcb.html
Result: 71cbce75031a driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Checks:
revert: PASS
verify: PASS

Parameters:
Tree: stable-rc
URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
Branch: linux-5.10.y
Target: r8a77960-ulcb
CPU arch: arm64
Lab: lab-collabora
Compiler: gcc-10
Config: defconfig
Test case: baseline.bootrr.deferred-probe-empty

Breaking commit found:

-------------------------------------------------------------------------------
commit 71cbce75031aed26c72c2dc8a83111d181685f1b
Author: Saravana Kannan <saravanak@...>
Date: Fri Jun 3 13:31:37 2022 +0200

driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] - https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] - https://lore.kernel.org/lkml/[email protected]/
[3] - https://lore.kernel.org/lkml/[email protected]/

Fixes 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires")
Cc John Stultz <jstultz@...>
Cc "David S. Miller" <davem@...>
Cc Alexey Kuznetsov <kuznet@...>
Cc Hideaki YOSHIFUJI <yoshfuji@...>
Cc Jakub Kicinski <kuba@...>
Cc Rob Herring <robh@...>
Cc Geert Uytterhoeven <geert@...>
Cc Yoshihiro Shimoda <yoshihiro.shimoda.uh@...>
Cc Robin Murphy <robin.murphy@...>
Cc Andy Shevchenko <andy.shevchenko@...>
Cc Sudeep Holla <sudeep.holla@...>
Cc Andy Shevchenko <andriy.shevchenko@...>
Cc Naresh Kamboju <naresh.kamboju@...>
Cc Basil Eljuse <Basil.Eljuse@...>
Cc Ferry Toth <fntoth@...>
Cc Arnd Bergmann <arnd@...>
Cc Anders Roxell <anders.roxell@...>
Cc linux-pm@...
drivers/base/dd.c +0 -0
index 4f4e8aedbd2c..f9d9f1ad9215 100644
@@ -250,7 +250,6 @@ DEFINE_SHOW_ATTRIBUTE(deferred_devs);
 

int driver_deferred_probe_timeout; EXPORT_SYMBOL_GPL(driver_deferred_probe_timeout); -static DECLARE_WAIT_QUEUE_HEAD(probe_timeout_waitqueue);

static int __init deferred_probe_timeout_setup(char *str)
{
@@ -302,7 +301,6 @@ static void deferred_probe_timeout_work_func(struct work_struct *work) list_for_each_entry(p, &deferred_probe_pending_list, deferred_probe) dev_info(p->device, "deferred probe pending\n"); mutex_unlock(&deferred_probe_mutex);
- wake_up_all(&probe_timeout_waitqueue);
}
static DECLARE_DELAYED_WORK(deferred_probe_timeout_work, deferred_probe_timeout_work_func);

@@ -706,9 +704,6 @@ int driver_probe_done(void)
*/
void wait_for_device_probe(void)
{
- /* wait for probe timeout */
- wait_event(probe_timeout_waitqueue, !driver_deferred_probe_timeout); -
/* wait for the deferred probe workqueue to finish */ flush_work(&deferred_probe_work);
-------------------------------------------------------------------------------

Git bisection log:

-------------------------------------------------------------------------------
git bisect start
# good: [2c85ebc57b3e1817b6ce1a6b703928e113a90442] Linux 5.10
git bisect good 2c85ebc57b3e1817b6ce1a6b703928e113a90442
# bad: [686c84f2f136412631eb684b064def993a96a8cc] Linux 5.10.189-rc1
git bisect bad 686c84f2f136412631eb684b064def993a96a8cc
# good: [88f1b613c37fbd3c4171f5a9decdcd12ae704637] Bluetooth: cmtp: fix possible panic when cmtp_init_sockets() fails git bisect good 88f1b613c37fbd3c4171f5a9decdcd12ae704637
# bad: [6c5742372b2d5d36de129439e26eda05aab54652] Input: snvs_pwrkey - fix SNVS_HPVIDR1 register address git bisect bad 6c5742372b2d5d36de129439e26eda05aab54652
# good: [07280d2c3f33d47741f42411eb8c976b70c6657a] random: make more consistent use of integer types
git bisect good 07280d2c3f33d47741f42411eb8c976b70c6657a
# bad: [2fc7f18ba2f98d15f174ce8e25a5afa46926eb55] tools headers: Remove broken definition of __LITTLE_ENDIAN git bisect bad 2fc7f18ba2f98d15f174ce8e25a5afa46926eb55
# bad: [c2ae49a113a5344232f1ebb93bcf18bbd11e9c39] net: dsa: lantiq_gswip: Fix refcount leak in gswip_gphy_fw_list git bisect bad c2ae49a113a5344232f1ebb93bcf18bbd11e9c39
# good: [c1b08aa568e829b743affe5d3231e6de28b7609e] ASoC: samsung: Use dev_err_probe() helper
git bisect good c1b08aa568e829b743affe5d3231e6de28b7609e
# good: [97a9ec86ccb4e336ecde46db42b59b2ff7e0d719] drm/nouveau/clk: Fix an incorrect NULL check on list iterator git bisect good 97a9ec86ccb4e336ecde46db42b59b2ff7e0d719
# good: [572211d631d7665c6690b5a6cb80436f8c368dc1] pwm: lp3943: Fix duty calculation in case period was clamped git bisect good 572211d631d7665c6690b5a6cb80436f8c368dc1
# good: [8f49e1694cbc29e76d5028267c1978cc2630e494] bpf: Fix probe read error in ___bpf_prog_run()
git bisect good 8f49e1694cbc29e76d5028267c1978cc2630e494
# bad: [3660db29b0305f9a1d95979c7af0f5db6ea99f5d] iommu/arm-smmu: fix possible null-ptr-deref in arm_smmu_device_probe() git bisect bad 3660db29b0305f9a1d95979c7af0f5db6ea99f5d
# good: [04622d631826ba483ae3a0b8a71c745d8e21453d] gpio: pca953x: use the correct register address to do regcache sync git bisect good 04622d631826ba483ae3a0b8a71c745d8e21453d
# bad: [32be2b805a1a13ccc68bd209ec3ae198dd3ba5d6] perf c2c: Fix sorting in percent_rmt_hitm_cmp()
git bisect bad 32be2b805a1a13ccc68bd209ec3ae198dd3ba5d6
# good: [c1f0187025905e9981000d44a92e159468b561a8] scsi: sd: Fix potential NULL pointer dereference
git bisect good c1f0187025905e9981000d44a92e159468b561a8
# bad: [71cbce75031aed26c72c2dc8a83111d181685f1b] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction git bisect bad 71cbce75031aed26c72c2dc8a83111d181685f1b
# good: [b8fac8e321044a9ac50f7185b4e9d91a7745e4b0] tipc: check attribute length for bearer name
git bisect good b8fac8e321044a9ac50f7185b4e9d91a7745e4b0

# first bad commit: [71cbce75031aed26c72c2dc8a83111d181685f1b] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction
-------------------------------------------------------------------------------
Quoted 63 lines
Fixes 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires")
Cc John Stultz<[email protected]>
Cc "David S. Miller"<[email protected]>
Cc Alexey Kuznetsov<[email protected]>
Cc Hideaki YOSHIFUJI<[email protected]>
Cc Jakub Kicinski<[email protected]>
Cc Rob Herring<[email protected]>
Cc Geert Uytterhoeven<[email protected]>
Cc Yoshihiro Shimoda<[email protected]>
Cc Robin Murphy<[email protected]>
Cc Andy Shevchenko<[email protected]>
Cc Sudeep Holla<[email protected]>
Cc Andy Shevchenko<[email protected]>
Cc Naresh Kamboju<[email protected]>
Cc Basil Eljuse<[email protected]>
Cc Ferry Toth<[email protected]>
Cc Arnd Bergmann<[email protected]>
Cc Anders Roxell<[email protected]>
---
  drivers/base/dd.c | 5 -----
  1 file changed, 5 deletions(-)
drivers/base/dd.c +0 -5
index 977e94cf669e..86fd2ea35656 100644
@@ -257,7 +257,6 @@ DEFINE_SHOW_ATTRIBUTE(deferred_devs);
 
  int driver_deferred_probe_timeout;
  EXPORT_SYMBOL_GPL(driver_deferred_probe_timeout);
-static DECLARE_WAIT_QUEUE_HEAD(probe_timeout_waitqueue);
 
  static int __init deferred_probe_timeout_setup(char *str)
  {
@@ -312,7 +311,6 @@ static void deferred_probe_timeout_work_func(struct work_struct *work)
  	list_for_each_entry(p, &deferred_probe_pending_list, deferred_probe)
  		dev_info(p->device, "deferred probe pending\n");
  	mutex_unlock(&deferred_probe_mutex);
-	wake_up_all(&probe_timeout_waitqueue);
  }
  static DECLARE_DELAYED_WORK(deferred_probe_timeout_work, deferred_probe_timeout_work_func);
 
@@ -720,9 +718,6 @@ int driver_probe_done(void)
   */
  void wait_for_device_probe(void)
  {
-	/* wait for probe timeout */
-	wait_event(probe_timeout_waitqueue, !driver_deferred_probe_timeout);
-
  	/* wait for the deferred probe workqueue to finish */
  	flush_work(&deferred_probe_work);
 
#129 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Hi Shreeya,

Quoted 36 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Quoted 5 lines

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

Commit 9be4cbd09da820a2 ("driver core: Set default deferred_probe_timeout back to 0.") in v5.19 contains a reference to the same commit as mentioned in the Fixes tag. Does backporting that help?

Gr{oetje,eeting}s,

Geert

Signature · 5 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds
#130 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 41 lines · Geert Uytterhoeven
Quoted 39 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

Quoted 6 lines
Quoted 2 lines

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

Commit 9be4cbd09da820a2 ("driver core: Set default deferred_probe_timeout back to 0.") in v5.19 contains a reference to the same commit as mentioned in the Fixes tag. Does backporting that help?

Anyway, remembering the days (weeks?) spent in investigating subtle issues with fw_devlinks and deferred probe, collecting all the fixes for backporting to stable may be a very hard job...

Gr{oetje,eeting}s,

Geert

Signature · 5 lines
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds
author maintainer
#131 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

In reply to Shreeya Patel
Quoted 100 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

confused,

greg k-h

#132 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Hi Greg,

Quoted 100 lines · Greg Kroah-Hartman
Quoted 99 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this issue, but this test does fail on all the stable releases from 5.10 onwards.

stable 5.15 :- https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :- https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Thanks,
Shreeya Patel

Quoted 5 lines

confused,

greg k-h

#133 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 110 lines · Shreeya Patel

Hi Greg,

Quoted 100 lines · Greg Kroah-Hartman
Quoted 99 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Quoted 9 lines

Sorry, I accidentally used the lore link for 5.17 while reporting this issue,
but this test does fail on all the stable releases from 5.10 onwards.

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

-Saravana

#134 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Hi Geert, Saravana,

Quoted 114 lines · Saravana Kannan
Quoted 112 lines · Shreeya Patel

Hi Greg,

Quoted 100 lines · Greg Kroah-Hartman
Quoted 99 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this issue,
but this test does fail on all the stable releases from 5.10 onwards.

Quoted 7 lines
Quoted 5 lines

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

I tried to test the kernel with 9be4cbd09da8 but it didn't change the result.
https://lava.collabora.dev/scheduler/job/11311615

Also, I am not sure if this can change things but just FYI, KernelCI adds some kernel parameters when running these tests and one of the parameter is deferred_probe_timeout=60. You can check this in the definition details given in the Lava job. I also tried to remove this parameter and rerun the test but again I got the same result.

I will try to add 9be4cbd09da8 to mainline kernel and see what results I get.

Thanks,
Shreeya Patel

Quoted 3 lines

-Saravana

#135 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 126 lines · Shreeya Patel

Hi Geert, Saravana,

Quoted 114 lines · Saravana Kannan
Quoted 112 lines · Shreeya Patel

Hi Greg,

Quoted 100 lines · Greg Kroah-Hartman
Quoted 99 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this issue,
but this test does fail on all the stable releases from 5.10 onwards.

Quoted 7 lines
Quoted 5 lines

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

Quoted 7 lines

I tried to test the kernel with 9be4cbd09da8 but it didn't change the result.
https://lava.collabora.dev/scheduler/job/11311615

Also, I am not sure if this can change things but just FYI, KernelCI adds some kernel parameters when running these tests and one of the parameter is deferred_probe_timeout=60.

Ah this is good to know.

Quoted 3 lines

You can check this in the definition details given in the Lava job. I also tried to remove this parameter and rerun the test but again I got the same result.

How long does the test wait after boot before checking for the deferred devices list?

Quoted 2 lines

I will try to add 9be4cbd09da8 to mainline kernel and see what results I get.

Now I'm confused. What do you mean by mainline? Are you saying the tip of tree of Linus's tree is also hitting this issue?

-Saravana

#136 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 133 lines · Saravana Kannan
Quoted 124 lines · Shreeya Patel

Hi Geert, Saravana,

Quoted 114 lines · Saravana Kannan
Quoted 112 lines · Shreeya Patel

Hi Greg,

Quoted 100 lines · Greg Kroah-Hartman
Quoted 99 lines · Shreeya Patel
Quoted 28 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1]. This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 6 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2] -https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this issue,
but this test does fail on all the stable releases from 5.10 onwards.

Quoted 7 lines
Quoted 5 lines

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

Quoted 7 lines

I tried to test the kernel with 9be4cbd09da8 but it didn't change the result.
https://lava.collabora.dev/scheduler/job/11311615

Also, I am not sure if this can change things but just FYI, KernelCI adds some kernel parameters when running these tests and one of the parameter is deferred_probe_timeout=60.

Quoted 8 lines

Ah this is good to know.

Quoted 3 lines

You can check this in the definition details given in the Lava job. I also tried to remove this parameter and rerun the test but again I got the same result.

How long does the test wait after boot before checking for the deferred devices list?

AFAIK, script for running the tests is immediately ran after the boot process is complete so there is no wait time.

Quoted 4 lines
Quoted 2 lines

I will try to add 9be4cbd09da8 to mainline kernel and see what results I get.

Now I'm confused. What do you mean by mainline? Are you saying the tip of tree of Linus's tree is also hitting this issue?

KernelCI runs tests on different kernel branches and trees, we also have this same test running on mainline tree. Following is the link to the dashboard for it and as you can see, it does fail there too.

https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Quoted 2 lines

-Saravana

#137 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 142 lines · Shreeya Patel
Quoted 140 lines · Saravana Kannan
Quoted 129 lines · Shreeya Patel

Hi Geert, Saravana,

Quoted 119 lines · Saravana Kannan
Quoted 117 lines · Shreeya Patel

Hi Greg,

Quoted 105 lines · Greg Kroah-Hartman
Quoted 104 lines · Shreeya Patel
Quoted 29 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1].  This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to
      hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Quoted 8 lines

Neither of this is good. So revert the changes to
wait_for_device_probe().

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2]
-https://lore.kernel.org/lkml/[email protected]/
[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case
<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this issue,
but this test does fail on all the stable releases from 5.10 onwards.

Quoted 7 lines
Quoted 5 lines

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

Quoted 7 lines

I tried to test the kernel with 9be4cbd09da8 but it didn't change the result.
https://lava.collabora.dev/scheduler/job/11311615

Also, I am not sure if this can change things but just FYI, KernelCI adds some kernel parameters when running these tests and one of the parameter is deferred_probe_timeout=60.

Ah this is good to know.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

You can check this in the definition details given in the Lava job. I also tried to remove this parameter and rerun the test but again I got the same result.

How long does the test wait after boot before checking for the deferred devices list?

AFAIK, script for running the tests is immediately ran after the boot process is complete so there is no wait time.

Regardless of what the kernel is doing, it seems like a fundamentally dumb test to specifically ask deferred probe to wait for up to a minute then complain that it hasn't finished after 11 seconds :/

If anything, it seems plausible that the "regression" might actually be the correct behaviour, and it was wrong before. I can't manage to pull up a boot log for a pre-5.10 kernel since all the async stuff on the KernelCI dashboard always just times out for me with a helpful "Error while loading data from the server (error code: 0)", but what would be interesting is whether those devices on the list are expected to successfully probe anyway - the mainline log below also shows other stuff failing to probe and CPUs failing to come online, so it's clearly not a very happy platform to begin with.

Robin.

Quoted 15 lines
Quoted 4 lines
Quoted 2 lines

I will try to add 9be4cbd09da8 to mainline kernel and see what results I get.

Now I'm confused. What do you mean by mainline? Are you saying the tip of tree of Linus's tree is also hitting this issue?

KernelCI runs tests on different kernel branches and trees, we also have this same test running on mainline tree. Following is the link to the dashboard for it and as you can see, it does fail there too.

https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Quoted 2 lines

-Saravana

#138 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 8 lines · Robin Murphy
Quoted 2 lines · Shreeya Patel

AFAIK, script for running the tests is immediately ran after the boot process is complete so there is no wait time.

Regardless of what the kernel is doing, it seems like a fundamentally dumb test to specifically ask deferred probe to wait for up to a minute then complain that it hasn't finished after 11 seconds :/

IIRC that stuff is expecting the modules to be loaded from the initramfs and checking from the main system which is a bit more sensible (at least in the case where there is a main filesystem). It's vulnerable to races but less so, especially given the time a Debian rootfs typically takes to boot over NFS.

#139 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

In reply to Robin Murphy

Hi Robin,

Quoted 153 lines · Robin Murphy
Quoted 152 lines · Shreeya Patel
Quoted 149 lines · Saravana Kannan
Quoted 132 lines · Shreeya Patel

Hi Geert, Saravana,

Quoted 122 lines · Saravana Kannan
Quoted 120 lines · Shreeya Patel

Hi Greg,

Quoted 108 lines · Greg Kroah-Hartman
Quoted 107 lines · Shreeya Patel
Quoted 33 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was non-zero [1].  This was because ip_auto_config() initcall times out waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred probe work or asynchronous probe finishes, that wasn't sufficient to account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires") tried to fix that by making sure wait_for_device_probe() waits for deferred_probe_timeout to expire before returning.

However, if wait_for_device_probe() is called from the kernel_init() context:

- Before deferred_probe_initcall() [2], it causes the boot process to
      hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from
      continuing till deferred_probe_timeout expires and beats the point of deferred_probe_timeout that's trying to wait for userspace to load modules.

Neither of this is good. So revert the changes to
wait_for_device_probe().

Quoted 6 lines

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2]
-https://lore.kernel.org/lkml/[email protected]/

[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb, see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case
<8>[   17.476741] <LAVA_SIGNAL_TESTCASE TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 5 lines
Quoted 2 lines

stable-rc/linux-5.10.y bisection: baseline.bootrr.deferred-probe-empty on r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this issue,
but this test does fail on all the stable releases from 5.10 onwards.

Quoted 7 lines
Quoted 5 lines

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

Quoted 7 lines

I tried to test the kernel with 9be4cbd09da8 but it didn't change the result.
https://lava.collabora.dev/scheduler/job/11311615

Also, I am not sure if this can change things but just FYI, KernelCI adds some kernel parameters when running these tests and one of the parameter is deferred_probe_timeout=60.

Ah this is good to know.

Quoted 3 lines

You can check this in the definition details given in the Lava job. I also tried to remove this parameter and rerun the test but again I got the same result.

How long does the test wait after boot before checking for the deferred devices list?

Quoted 6 lines
Quoted 1 line

AFAIK, script for running the tests is immediately ran after the boot process is complete so there is no wait time.

Regardless of what the kernel is doing, it seems like a fundamentally dumb test to specifically ask deferred probe to wait for up to a minute then complain that it hasn't finished after 11 seconds :/

If anything, it seems plausible that the "regression" might actually be the correct behaviour, and it was wrong before. I can't manage to pull up a boot log for a pre-5.10 kernel since all the async stuff on the KernelCI dashboard always just times out for me with a helpful "Error while loading data from the server (error code: 0)", but what would be interesting is whether those devices on the list are expected to successfully probe anyway - the mainline log below also shows other stuff failing to probe and CPUs failing to come online, so it's clearly not a very happy platform to begin with.

Sorry about the dashboard issues you are facing, KernelCI team is working on a new dashboard which will fix all of these issues. But we need to wait for it be ready for some more time.

Your point makes sense and that is why we did a test to add some sleep time of 60-65 seconds before running the tests and it actually fixed the problem. There are no more deferred devices as you can see in the following job.
https://lava.collabora.dev/scheduler/job/11330931

This change, to add deferred_probe_timeout=60 as kernel parameter was recently added in KernelCI since there were number of devices failing to probe on different platforms, specifically for chromebooks. Unfortunately, no one realized to change the start of the test time and hence these issues are seen now on different platforms.

Thanks for pointing it out, it will help us to eliminate quite many deferred probe test failures that we are seeing.

Saravana, thanks to you as well for asking the valid questions about when the test starts running.

Thanks,
Shreeya Patel

#regzbot resolve: Nothing is broken due to the patch, a fix is needed on the KernelCI side.

Quoted 18 lines

Robin.

Quoted 15 lines
Quoted 4 lines
Quoted 2 lines

I will try to add 9be4cbd09da8 to mainline kernel and see what results I get.

Now I'm confused. What do you mean by mainline? Are you saying the tip of tree of Linus's tree is also hitting this issue?

KernelCI runs tests on different kernel branches and trees, we also have this same test running on mainline tree. Following is the link to the dashboard for it and as you can see, it does fail there too.

https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Quoted 2 lines

-Saravana

#140 lore raw

Re: [PATCH 5.17 127/298] driver core: Fix wait_for_device_probe() & deferred_probe_timeout interaction

Quoted 210 lines · Shreeya Patel

Hi Robin,

Quoted 178 lines · Robin Murphy
Quoted 166 lines · Shreeya Patel
Quoted 164 lines · Saravana Kannan
Quoted 150 lines · Shreeya Patel

Hi Geert, Saravana,

Quoted 140 lines · Saravana Kannan
Quoted 138 lines · Shreeya Patel

Hi Greg,

Quoted 123 lines · Greg Kroah-Hartman
Quoted 122 lines · Shreeya Patel
Quoted 43 lines · Greg Kroah-Hartman

From: Saravana Kannan<[email protected]>

[ Upstream commit 5ee76c256e928455212ab759c51d198fedbe7523 ]

Mounting NFS rootfs was timing out when deferred_probe_timeout was
non-zero [1]. This was because ip_auto_config() initcall times out
waiting for the network interfaces to show up when deferred_probe_timeout was non-zero. While ip_auto_config() calls wait_for_device_probe() to make sure any currently running deferred
probe work or asynchronous probe finishes, that wasn't sufficient to
account for devices being deferred until deferred_probe_timeout.

Commit 35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits
until the deferred_probe_timeout fires") tried to fix that by making
sure wait_for_device_probe() waits for deferred_probe_timeout to expire
before returning.

However, if wait_for_device_probe() is called from the kernel_init()
context:

- Before deferred_probe_initcall() [2], it causes the boot process to
hang due to a deadlock.

- After deferred_probe_initcall() [3], it blocks kernel_init() from
continuing till deferred_probe_timeout expires and beats
the point of
deferred_probe_timeout that's trying to wait for
userspace to load
modules.

Neither of this is good. So revert the changes to
wait_for_device_probe().

Quoted 6 lines

[1] -https://lore.kernel.org/lkml/TYAPR01MB45443DF63B9EF29054F7C41FD8C60@TYAPR01MB4544.jpnprd01.prod.outlook.com/
[2]
-https://lore.kernel.org/lkml/[email protected]/

[3] -https://lore.kernel.org/lkml/[email protected]/

Hi Saravana, Greg,

KernelCI found this patch causes the baseline.bootrr.deferred-probe-empty test to fail on r8a77960-ulcb,
see the following details for more information.

KernelCI dashboard link:
https://linux.kernelci.org/test/plan/id/64d2a6be8c1a8435e535b264/

Error messages from the logs :-

+ UUID=11236495_1.5.2.4.5
+ set +x
+ export 'PATH=/opt/bootrr/libexec/bootrr/helpers:/lava-11236495/1/../bin:/sbin:/usr/sbin:/bin:/usr/bin'
+ cd /opt/bootrr/libexec/bootrr
+ sh helpers/bootrr-auto
e6800000.ethernet
e6700000.dma-controller
e7300000.dma-controller
e7310000.dma-controller
ec700000.dma-controller
ec720000.dma-controller
fea20000.vsp
feb00000.display
fea28000.vsp
fea30000.vsp
fe9a0000.vsp
fe9af000.fcp
fea27000.fcp
fea2f000.fcp
fea37000.fcp
sound
ee100000.mmc
ee140000.mmc
ec500000.sound
/lava-11236495/1/../bin/lava-test-case

<8>[   17.476741] <LAVA_SIGNAL_TESTCASE

TEST_CASE_ID=deferred-probe-empty RESULT=fail>

Test case failing :-
Baseline Bootrr deferred-probe-empty test -https://github.com/kernelci/bootrr/blob/main/helpers/bootrr-generic-tests

Regression Reproduced :-

Lava job after reverting the commit 5ee76c256e92
https://lava.collabora.dev/scheduler/job/11292890

Bisection report from KernelCI can be found at the bottom of the email.

Thanks,
Shreeya Patel

#regzbot introduced: 5ee76c256e92
#regzbot title: KernelCI: Multiple devices deferring on r8a77960-ulcb

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

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **
* If you do send a fix, please include this trailer: *
* Reported-by: "kernelci.org bot" <bot@...> *
* *
* Hope this helps! *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Quoted 6 lines
Quoted 3 lines

stable-rc/linux-5.10.y bisection:
baseline.bootrr.deferred-probe-empty on
r8a77960-ulcb

You are testing 5.10.y, yet the subject says 5.17?

Which is it here?

Sorry, I accidentally used the lore link for 5.17 while reporting this
issue,
but this test does fail on all the stable releases from 5.10 onwards.

Quoted 7 lines
Quoted 5 lines

stable 5.15 :-
https://linux.kernelci.org/test/case/id/64dd156a5ac58d0cf335b1ea/
mainline :-
https://linux.kernelci.org/test/case/id/64dc13d55cb51357a135b209/

Shreeya, can you try the patch Geert suggested and let us know if it helps? If not, then I can try to take a closer look.

Quoted 7 lines

I tried to test the kernel with 9be4cbd09da8 but it didn't change the result.
https://lava.collabora.dev/scheduler/job/11311615

Also, I am not sure if this can change things but just FYI, KernelCI adds some kernel parameters when running these tests and one of the parameter is deferred_probe_timeout=60.

Ah this is good to know.

Quoted 3 lines

You can check this in the definition details given in the Lava job. I also tried to remove this parameter and rerun the test but again I got the same result.

Quoted 6 lines
Quoted 3 lines

How long does the test wait after boot before checking for the deferred devices list?

AFAIK, script for running the tests is immediately ran after the boot process is complete so there is no wait time.

Regardless of what the kernel is doing, it seems like a fundamentally dumb test to specifically ask deferred probe to wait for up to a minute then complain that it hasn't finished after 11 seconds :/

Quoted 10 lines

If anything, it seems plausible that the "regression" might actually be the correct behaviour, and it was wrong before. I can't manage to pull up a boot log for a pre-5.10 kernel since all the async stuff on the KernelCI dashboard always just times out for me with a helpful "Error while loading data from the server (error code: 0)", but what would be interesting is whether those devices on the list are expected to successfully probe anyway - the mainline log below also shows other stuff failing to probe and CPUs failing to come online, so it's clearly not a very happy platform to begin with.

Sorry about the dashboard issues you are facing, KernelCI team is working on a new dashboard which will fix all of these issues. But we need to wait for it be ready for some more time.

Your point makes sense and that is why we did a test to add some sleep time of 60-65 seconds before running the tests and it actually fixed the problem. There are no more deferred devices as you can see in the following job.
https://lava.collabora.dev/scheduler/job/11330931

This change, to add deferred_probe_timeout=60 as kernel parameter was recently added in KernelCI since there were number of devices failing to probe on different platforms, specifically for chromebooks. Unfortunately, no one realized to change the start of the test time and hence these issues are seen now on different platforms.

Quoted 5 lines

Thanks for pointing it out, it will help us to eliminate quite many deferred probe test failures that we are seeing.

Saravana, thanks to you as well for asking the valid questions about when the test starts running.

Glad we sorted this out. Thanks for maintaining the tests.

-Saravana

author maintainer
#141 lore raw

PATCH 128/298 perf evsel: Fixes topdown events in a weak group for the hybrid platform

In reply to Greg Kroah-Hartman

From: Kan Liang <[email protected]>

[ Upstream commit 39d5f412da84784bcc7f39ed49e55376be526fc7 ]

The patch ("perf evlist: Keep topdown counters in weak group") fixes the perf metrics topdown event issue when the topdown events are in a weak group on a non-hybrid platform. However, it doesn't work for the hybrid platform.

$./perf stat -e '{cpu_core/slots/,cpu_core/topdown-bad-spec/, cpu_core/topdown-be-bound/,cpu_core/topdown-fe-bound/, cpu_core/topdown-retiring/,cpu_core/branch-instructions/, cpu_core/branch-misses/,cpu_core/bus-cycles/,cpu_core/cache-misses/, cpu_core/cache-references/,cpu_core/cpu-cycles/,cpu_core/instructions/, cpu_core/mem-loads/,cpu_core/mem-stores/,cpu_core/ref-cycles/, cpu_core/cache-misses/,cpu_core/cache-references/}:W' -a sleep 1

Performance counter stats for 'system wide':

       751,765,068      cpu_core/slots/                        (84.07%)
   <not supported>      cpu_core/topdown-bad-spec/
   <not supported>      cpu_core/topdown-be-bound/
   <not supported>      cpu_core/topdown-fe-bound/
   <not supported>      cpu_core/topdown-retiring/
        12,398,197      cpu_core/branch-instructions/          (84.07%)
         1,054,218      cpu_core/branch-misses/                (84.24%)
       539,764,637      cpu_core/bus-cycles/                   (84.64%)
            14,683      cpu_core/cache-misses/                 (84.87%)
         7,277,809      cpu_core/cache-references/             (77.30%)
       222,299,439      cpu_core/cpu-cycles/                   (77.28%)
        63,661,714      cpu_core/instructions/                 (84.85%)
                 0      cpu_core/mem-loads/                    (77.29%)
        12,271,725      cpu_core/mem-stores/                   (77.30%)
       542,241,102      cpu_core/ref-cycles/                   (84.85%)
             8,854      cpu_core/cache-misses/                 (76.71%)
         7,179,013      cpu_core/cache-references/             (76.31%)
         1.003245250 seconds time elapsed

A hybrid platform has a different PMU name for the core PMUs, while the current perf hard code the PMU name "cpu".

The evsel->pmu_name can be used to replace the "cpu" to fix the issue. For a hybrid platform, the pmu_name must be non-NULL. Because there are at least two core PMUs. The PMU has to be specified.
For a non-hybrid platform, the pmu_name may be NULL. Because there is only one core PMU, "cpu". For a NULL pmu_name, we can safely assume that it is a "cpu" PMU.

In case other PMUs also define the "slots" event, checking the PMU type as well.

With the patch,

  $ perf stat -e '{cpu_core/slots/,cpu_core/topdown-bad-spec/,
  cpu_core/topdown-be-bound/,cpu_core/topdown-fe-bound/,
  cpu_core/topdown-retiring/,cpu_core/branch-instructions/,
  cpu_core/branch-misses/,cpu_core/bus-cycles/,cpu_core/cache-misses/,
  cpu_core/cache-references/,cpu_core/cpu-cycles/,cpu_core/instructions/,
  cpu_core/mem-loads/,cpu_core/mem-stores/,cpu_core/ref-cycles/,
  cpu_core/cache-misses/,cpu_core/cache-references/}:W' -a sleep 1

Performance counter stats for 'system wide':

     766,620,266   cpu_core/slots/                                        (84.06%)
      73,172,129   cpu_core/topdown-bad-spec/ #    9.5% bad speculation   (84.06%)
     193,443,341   cpu_core/topdown-be-bound/ #    25.0% backend bound    (84.06%)
     403,940,929   cpu_core/topdown-fe-bound/ #    52.3% frontend bound   (84.06%)
     102,070,237   cpu_core/topdown-retiring/ #    13.2% retiring         (84.06%)
      12,364,429   cpu_core/branch-instructions/                          (84.03%)
       1,080,124   cpu_core/branch-misses/                                (84.24%)
     564,120,383   cpu_core/bus-cycles/                                   (84.65%)
          36,979   cpu_core/cache-misses/                                 (84.86%)
       7,298,094   cpu_core/cache-references/                             (77.30%)
     227,174,372   cpu_core/cpu-cycles/                                   (77.31%)
      63,886,523   cpu_core/instructions/                                 (84.87%)
               0   cpu_core/mem-loads/                                    (77.31%)
      12,208,782   cpu_core/mem-stores/                                   (77.31%)
     566,409,738   cpu_core/ref-cycles/                                   (84.87%)
          23,118   cpu_core/cache-misses/                                 (76.71%)
       7,212,602   cpu_core/cache-references/                             (76.29%)
       1.003228667 seconds time elapsed
---
 tools/perf/arch/x86/util/evsel.c | 23 +++++++++++++++++++++--
 1 file changed, 21 insertions(+), 2 deletions(-)
tools/perf/arch/x86/util/evsel.c +21 -2
index 0c9e56ab07b5..ff4561b7b600 100644
@@ -31,10 +31,29 @@ void arch_evsel__fixup_new_cycles(struct perf_event_attr *attr)
 	free(env.cpuid);
 }
 
+/* Check whether the evsel's PMU supports the perf metrics */
+static bool evsel__sys_has_perf_metrics(const struct evsel *evsel)
+{
+	const char *pmu_name = evsel->pmu_name ? evsel->pmu_name : "cpu";
+
+	/*
+	 * The PERF_TYPE_RAW type is the core PMU type, e.g., "cpu" PMU
+	 * on a non-hybrid machine, "cpu_core" PMU on a hybrid machine.
+	 * The slots event is only available for the core PMU, which
+	 * supports the perf metrics feature.
+	 * Checking both the PERF_TYPE_RAW type and the slots event
+	 * should be good enough to detect the perf metrics feature.
+	 */
+	if ((evsel->core.attr.type == PERF_TYPE_RAW) &&
+	    pmu_have_event(pmu_name, "slots"))
+		return true;
+
+	return false;
+}
+
 bool arch_evsel__must_be_in_group(const struct evsel *evsel)
 {
-	if ((evsel->pmu_name && strcmp(evsel->pmu_name, "cpu")) ||
-	    !pmu_have_event("cpu", "slots"))
+	if (!evsel__sys_has_perf_metrics(evsel))
 		return false;
 
 	return evsel->name &&
Signature · 1 line
2.35.1
author maintainer
#142 lore raw

PATCH 129/298 perf parse-events: Move slots event for the hybrid platform too

In reply to Greg Kroah-Hartman

From: Kan Liang <[email protected]>

[ Upstream commit e0e14cdff31d326f81e0edbd5140f788c870756c ]

The commit 94dbfd6781a0e87b ("perf parse-events: Architecture specific leader override") introduced a feature to reorder the slots event to fulfill the restriction of the perf metrics topdown group. But the feature doesn't work on the hybrid machine.

  $ perf stat -e "{cpu_core/instructions/,cpu_core/slots/,cpu_core/topdown-retiring/}" -a sleep 1

Performance counter stats for 'system wide':

       <not counted>      cpu_core/instructions/
       <not counted>      cpu_core/slots/
     <not supported>      cpu_core/topdown-retiring/
         1.002871801 seconds time elapsed

A hybrid platform has a different PMU name for the core PMUs, while current perf hard code the PMU name "cpu".

Introduce a new function to check whether the system supports the perf metrics feature. The result is cached for the future usage.

For X86, the core PMU name always has "cpu" prefix.

With the patch:

  $ perf stat -e "{cpu_core/instructions/,cpu_core/slots/,cpu_core/topdown-retiring/}" -a sleep 1

Performance counter stats for 'system wide':

          76,337,010      cpu_core/slots/
          10,416,809      cpu_core/instructions/
          11,692,372      cpu_core/topdown-retiring/
         1.002805453 seconds time elapsed
---
 tools/perf/arch/x86/util/evlist.c  |  5 +++--
 tools/perf/arch/x86/util/topdown.c | 25 +++++++++++++++++++++++++
 tools/perf/arch/x86/util/topdown.h |  7 +++++++
 3 files changed, 35 insertions(+), 2 deletions(-)
 create mode 100644 tools/perf/arch/x86/util/topdown.h
tools/perf/arch/x86/util/evlist.c +3 -2
index 75564a7df15b..68f681ad54c1 100644
@@ -3,6 +3,7 @@
 #include "util/pmu.h"
 #include "util/evlist.h"
 #include "util/parse-events.h"
+#include "topdown.h"
 
 #define TOPDOWN_L1_EVENTS	"{slots,topdown-retiring,topdown-bad-spec,topdown-fe-bound,topdown-be-bound}"
 #define TOPDOWN_L2_EVENTS	"{slots,topdown-retiring,topdown-bad-spec,topdown-fe-bound,topdown-be-bound,topdown-heavy-ops,topdown-br-mispredict,topdown-fetch-lat,topdown-mem-bound}"
@@ -25,12 +26,12 @@ struct evsel *arch_evlist__leader(struct list_head *list)
 
 	first = list_first_entry(list, struct evsel, core.node);
 
-	if (!pmu_have_event("cpu", "slots"))
+	if (!topdown_sys_has_perf_metrics())
 		return first;
 
 	/* If there is a slots event and a topdown event then the slots event comes first. */
 	__evlist__for_each_entry(list, evsel) {
-		if (evsel->pmu_name && !strcmp(evsel->pmu_name, "cpu") && evsel->name) {
+		if (evsel->pmu_name && !strncmp(evsel->pmu_name, "cpu", 3) && evsel->name) {
 			if (strcasestr(evsel->name, "slots")) {
 				slots = evsel;
 				if (slots == first)
tools/perf/arch/x86/util/topdown.c +25 -0
index 2f3d96aa92a5..f4d5422e9960 100644
@@ -3,6 +3,31 @@
 #include "api/fs/fs.h"
 #include "util/pmu.h"
 #include "util/topdown.h"
+#include "topdown.h"
+
+/* Check whether there is a PMU which supports the perf metrics. */
+bool topdown_sys_has_perf_metrics(void)
+{
+	static bool has_perf_metrics;
+	static bool cached;
+	struct perf_pmu *pmu;
+
+	if (cached)
+		return has_perf_metrics;
+
+	/*
+	 * The perf metrics feature is a core PMU feature.
+	 * The PERF_TYPE_RAW type is the type of a core PMU.
+	 * The slots event is only available when the core PMU
+	 * supports the perf metrics feature.
+	 */
+	pmu = perf_pmu__find_by_type(PERF_TYPE_RAW);
+	if (pmu && pmu_have_event(pmu->name, "slots"))
+		has_perf_metrics = true;
+
+	cached = true;
+	return has_perf_metrics;
+}
 
 /*
  * Check whether we can use a group for top down.
tools/perf/arch/x86/util/topdown.h new file +7 -0
new file mode 100644
index 000000000000..46bf9273e572
@@ -0,0 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _TOPDOWN_H
+#define _TOPDOWN_H 1
+
+bool topdown_sys_has_perf_metrics(void);
+
+#endif
Signature · 1 line
2.35.1
author maintainer
#143 lore raw

PATCH 130/298 perf record: Support sample-read topdown metric group for hybrid platforms

In reply to Greg Kroah-Hartman

From: Zhengjun Xing <[email protected]>

[ Upstream commit 151e7d75036b4e2ac0f33730bc1a5b3ff424d9a7 ]

With the hardware TopDown metrics feature, the sample-read feature should be supported for a TopDown group, e.g., sample a non-topdown event and read a Topdown metric group. But the current perf record code errors are out.

For a TopDown metric group,the slots event must be the leader of the group, but the leader slots event doesn't support sampling. To support sample-read the TopDown metric group, uses the 2nd event of the group as the "leader" for the purposes of sampling.

Only the platform with the TopDown metric feature supports sample-read the topdown group. In commit acb65150a47c ("perf record: Support sample-read topdown metric group"), it adds arch_topdown_sample_read() to indicate whether the TopDown group supports sample-read, it should only work on the non-hybrid systems, this patch extends the support for hybrid platforms.

Before:

  # ./perf record -e "{cpu_core/slots/,cpu_core/cycles/,cpu_core/topdown-retiring/}:S" -a sleep 1
  Error:
  The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (cpu_core/topdown-retiring/).
  /bin/dmesg | grep -i perf may provide additional information.

After:

  # ./perf record -e "{cpu_core/slots/,cpu_core/cycles/,cpu_core/topdown-retiring/}:S" -a sleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.238 MB perf.data (369 samples) ]
Fixes acb65150a47c2bae ("perf record: Support sample-read topdown metric group")
Cc Adrian Hunter <[email protected]>
Cc Alexander Shishkin <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc Peter Zijlstra <[email protected]>
---
 tools/perf/arch/x86/util/evsel.c   |  3 ++-
 tools/perf/arch/x86/util/evsel.h   |  7 +++++++
 tools/perf/arch/x86/util/topdown.c | 21 ++++-----------------
 3 files changed, 13 insertions(+), 18 deletions(-)
 create mode 100644 tools/perf/arch/x86/util/evsel.h
tools/perf/arch/x86/util/evsel.c +2 -1
index ff4561b7b600..3501399cef35 100644
@@ -5,6 +5,7 @@
 #include "util/env.h"
 #include "util/pmu.h"
 #include "linux/string.h"
+#include "evsel.h"
 
 void arch_evsel__set_sample_weight(struct evsel *evsel)
 {
@@ -32,7 +33,7 @@ void arch_evsel__fixup_new_cycles(struct perf_event_attr *attr)
 }
 
 /* Check whether the evsel's PMU supports the perf metrics */
-static bool evsel__sys_has_perf_metrics(const struct evsel *evsel)
+bool evsel__sys_has_perf_metrics(const struct evsel *evsel)
 {
 	const char *pmu_name = evsel->pmu_name ? evsel->pmu_name : "cpu";
 
tools/perf/arch/x86/util/evsel.h new file +7 -0
new file mode 100644
index 000000000000..19ad1691374d
@@ -0,0 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _EVSEL_H
+#define _EVSEL_H 1
+
+bool evsel__sys_has_perf_metrics(const struct evsel *evsel);
+
+#endif
tools/perf/arch/x86/util/topdown.c +4 -17
index f4d5422e9960..f81a7cfe4d63 100644
@@ -4,6 +4,7 @@
 #include "util/pmu.h"
 #include "util/topdown.h"
 #include "topdown.h"
+#include "evsel.h"
 
 /* Check whether there is a PMU which supports the perf metrics. */
 bool topdown_sys_has_perf_metrics(void)
@@ -55,33 +56,19 @@ void arch_topdown_group_warn(void)
 
 #define TOPDOWN_SLOTS		0x0400
 
-static bool is_topdown_slots_event(struct evsel *counter)
-{
-	if (!counter->pmu_name)
-		return false;
-
-	if (strcmp(counter->pmu_name, "cpu"))
-		return false;
-
-	if (counter->core.attr.config == TOPDOWN_SLOTS)
-		return true;
-
-	return false;
-}
-
 /*
  * Check whether a topdown group supports sample-read.
  *
- * Only Topdown metic supports sample-read. The slots
+ * Only Topdown metric supports sample-read. The slots
  * event must be the leader of the topdown group.
  */
 
 bool arch_topdown_sample_read(struct evsel *leader)
 {
-	if (!pmu_have_event("cpu", "slots"))
+	if (!evsel__sys_has_perf_metrics(leader))
 		return false;
 
-	if (is_topdown_slots_event(leader))
+	if (leader->core.attr.config == TOPDOWN_SLOTS)
 		return true;
 
 	return false;
Signature · 1 line
2.35.1
author maintainer
#144 lore raw

PATCH 131/298 perf c2c: Fix sorting in percent_rmt_hitm_cmp()

In reply to Greg Kroah-Hartman

From: Leo Yan <[email protected]>

[ Upstream commit b24192a17337abbf3f44aaa75e15df14a2d0016e ]

The function percent_rmt_hitm_cmp() wrongly uses local HITMs for sorting remote HITMs.

Since this function is to sort cache lines for remote HITMs, this patch changes to use 'rmt_hitm' field for correct sorting.

Fixes 9cb3500afc0980c5 ("perf c2c report: Add hitm/store percent related sort keys")
Cc Alexander Shishkin <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc Joe Mario <[email protected]>
Cc Mark Rutland <[email protected]>
Cc Peter Zijlstra <[email protected]>
---
 tools/perf/builtin-c2c.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
tools/perf/builtin-c2c.c +2 -2
index d8ec683b06a5..4e0c385427a4 100644
@@ -924,8 +924,8 @@ percent_rmt_hitm_cmp(struct perf_hpp_fmt *fmt __maybe_unused,
 	double per_left;
 	double per_right;
 
-	per_left  = PERCENT(left, lcl_hitm);
-	per_right = PERCENT(right, lcl_hitm);
+	per_left  = PERCENT(left, rmt_hitm);
+	per_right = PERCENT(right, rmt_hitm);
 
 	return per_left - per_right;
 }
Signature · 1 line
2.35.1