* refs/heads/tmp-de197c5:
Linux 5.4.8
mm/hugetlbfs: fix for_each_hstate() loop in init_hugetlbfs_fs()
mmc: sdhci-of-esdhc: re-implement erratum A-009204 workaround
mmc: sdhci-of-esdhc: fix up erratum A-008171 workaround
vhost/vsock: accept only packets with the right dst_cid
net: ena: fix napi handler misbehavior when the napi budget is zero
net: phylink: fix interface passed to mac_link_up
ipv6/addrconf: only check invalid header values when NETLINK_F_STRICT_CHK is set
bnxt: apply computed clamp value for coalece parameter
gtp: do not allow adding duplicate tid and ms_addr pdp context
gtp: fix an use-after-free in ipv4_pdp_find()
hv_netvsc: Fix tx_table init in rndis_set_subchannel()
tcp/dccp: fix possible race __inet_lookup_established()
tcp: do not send empty skb from tcp_write_xmit()
bonding: fix active-backup transition after link failure
gtp: avoid zero size hashtable
gtp: fix wrong condition in gtp_genl_dump_pdp()
net: marvell: mvpp2: phylink requires the link interrupt
net: dsa: sja1105: Reconcile the meaning of TPID and TPID2 for E/T and P/Q/R/S
net/dst: do not confirm neighbor for vxlan and geneve pmtu update
sit: do not confirm neighbor when do pmtu update
vti: do not confirm neighbor when do pmtu update
tunnel: do not confirm neighbor when do pmtu update
net/dst: add new function skb_dst_update_pmtu_no_confirm
gtp: do not confirm neighbor when do pmtu update
ip6_gre: do not confirm neighbor when do pmtu update
net: add bool confirm_neigh parameter for dst_ops.update_pmtu
mlxsw: spectrum: Use dedicated policer for VRRP packets
mlxsw: spectrum_router: Skip loopback RIFs during MAC validation
bnxt_en: Add missing devlink health reporters for VFs.
bnxt_en: Fix the logic that creates the health reporters.
bnxt_en: Remove unnecessary NULL checks for fw_health
bnxt_en: Fix bp->fw_health allocation and free logic.
bnxt_en: Return error if FW returns more data than dump length
bnxt_en: Free context memory in the open path if firmware has been reset.
bnxt_en: Fix MSIX request logic for RDMA driver.
udp: fix integer overflow while computing available space in sk_rcvbuf
tcp: Fix highest_sack and highest_sack_seq
ptp: fix the race between the release of ptp_clock and cdev
net: stmmac: dwmac-meson8b: Fix the RGMII TX delay on Meson8b/8m2 SoCs
net_sched: sch_fq: properly set sk->sk_pacing_status
net/sched: add delete_empty() to filters and use it in cls_flower
net/sched: act_mirred: Pull mac prior redir to non mac_header_xmit device
net: phy: aquantia: add suspend / resume ops for AQR105
net/mlxfw: Fix out-of-memory error in mfa2 flash burning
net: dsa: bcm_sf2: Fix IP fragment location and behavior
cxgb4/cxgb4vf: fix flow control display for auto negotiation
xfs: fix mount failure crash on invalid iclog memory access
drm: limit to INT_MAX in create_blob ioctl
uaccess: disallow > INT_MAX copy sizes
tomoyo: Don't use nifty names on sockets.
hrtimer: Annotate lockless access to timer->state
net: icmp: fix data-race in cmp_global_allow()
net: add a READ_ONCE() in skb_peek_tail()
inetpeer: fix data-race in inet_putpeer / inet_putpeer
netfilter: bridge: make sure to pull arp header in br_nf_forward_arp()
net/smc: add fallback check to connect()
powerpc: Fix __clear_user() with KUAP enabled
6pack,mkiss: fix possible deadlock
netfilter: ebtables: compat: reject all padding in matches/watchers
Revert "iwlwifi: assign directly to iwl_trans->cfg in QuZ detection"
md: make sure desc_nr less than MD_SB_DISKS
sctp: fix err handling of stream initialization
Revert "powerpc/vcpu: Assume dedicated processors as non-preempt"
userfaultfd: require CAP_SYS_PTRACE for UFFD_FEATURE_EVENT_FORK
kernel: sysctl: make drop_caches write-only
mm/hugetlbfs: fix error handling when setting up mounts
selftests: vm: add fragment CONFIG_TEST_VMALLOC
s390: disable preemption when switching to nodat stack with CALL_ON_STACK
mailbox: imx: Fix Tx doorbell shutdown path
ocfs2: fix passing zero to 'PTR_ERR' warning
s390/cpum_sf: Check for SDBT and SDB consistency
s390/unwind: filter out unreliable bogus %r14
libfdt: define INT32_MAX and UINT32_MAX in libfdt_env.h
mailbox: imx: Clear the right interrupts at shutdown
s390/zcrypt: handle new reply code FILTERED_BY_HYPERVISOR
perf regs: Make perf_reg_name() return "unknown" instead of NULL
perf script: Fix brstackinsn for AUXTRACE
perf diff: Use llabs() with 64-bit values
cifs: move cifsFileInfo_put logic into a work-queue
cdrom: respect device capabilities during opening action
of: unittest: fix memory leak in attach_node_and_children
io_uring: io_allocate_scq_urings() should return a sane state
um: virtio: Keep reading on -EAGAIN
cifs: Fix use-after-free bug in cifs_reconnect()
powerpc: Don't add -mabi= flags when building with Clang
scripts/kallsyms: fix definitely-lost memory leak
drm/amdgpu: Call find_vma under mmap_sem
apparmor: fix unsigned len comparison with less than zero
Drivers: hv: vmbus: Fix crash handler reset of Hyper-V synic
tools/power/x86/intel-speed-select: Ignore missing config level
gpio: lynxpoint: Setup correct IRQ handlers
gpio: mpc8xxx: Don't overwrite default irq_set_type callback
platform/x86: intel_pmc_core: Add Comet Lake (CML) platform support to intel_pmc_core driver
platform/x86: intel_pmc_core: Fix the SoC naming inconsistency
gpio/mpc8xxx: fix qoriq GPIO reading
habanalabs: skip VA block list update in reset flow
f2fs: Fix deadlock in f2fs_gc() context during atomic files handling
scsi: target: iscsi: Wait for all commands to finish before freeing a session
scsi: iscsi: Don't send data to unbound connection
scsi: ufs: Fix up auto hibern8 enablement
scsi: target: core: Release SPC-2 reservations when closing a session
scsi: NCR5380: Add disconnect_mask module parameter
scsi: scsi_debug: num_tgts must be >= 0
scsi: ufs: Fix error handing during hibern8 enter
scsi: pm80xx: Fix for SATA device discovery
powerpc/fixmap: Use __fix_to_virt() instead of fix_to_virt()
watchdog: Fix the race between the release of watchdog_core_data and cdev
watchdog: prevent deferral of watchdogd wakeup on RT
watchdog: imx7ulp: Fix reboot hang
HID: rmi: Check that the RMI_STARTED bit is set before unregistering the RMI transport device
HID: Improve Windows Precision Touchpad detection.
libnvdimm/btt: fix variable 'rc' set but not used
ARM: 8937/1: spectre-v2: remove Brahma-B53 from hardening
HID: i2c-hid: fix no irq after reset on raydium 3118
HID: logitech-hidpp: Silence intermittent get_battery_capacity errors
dt-bindings: Improve validation build error handling
HID: quirks: Add quirk for HP MSU1465 PIXART OEM mouse
bcache: at least try to shrink 1 node in bch_mca_scan()
clk: pxa: fix one of the pxa RTC clocks
scsi: atari_scsi: sun3_scsi: Set sg_tablesize to 1 instead of SG_NONE
powerpc/book3s/mm: Update Oops message to print the correct translation in use
powerpc/eeh: differentiate duplicate detection message
powerpc/security: Fix wrong message when RFI Flush is disable
PCI: rpaphp: Correctly match ibm, my-drc-index to drc-name when using drc-info
PCI: rpaphp: Annotate and correctly byte swap DRC properties
PCI: rpaphp: Don't rely on firmware feature to imply drc-info support
powerpc/pseries/cmm: Implement release() function for sysfs device
scsi: ufs: fix potential bug which ends in system hang
PCI: rpaphp: Fix up pointer to first drc-info entry
scsi: zorro_esp: Limit DMA transfers to 65536 bytes (except on Fastlane)
scsi: lpfc: fix: Coverity: lpfc_cmpl_els_rsp(): Null pointer dereferences
Input: ili210x - handle errors from input_mt_init_slots()
iomap: fix return value of iomap_dio_bio_actor on 32bit systems
i2c: stm32f7: fix & reorder remove & probe error handling
iommu/arm-smmu-v3: Don't display an error when IRQ lines are missing
fs/quota: handle overflows of sysctl fs.quota.* and report as unsigned long
dma-direct: check for overflows on 32 bit DMA addresses
irqchip: ingenic: Error out if IRQ domain creation failed
irqchip/irq-bcm7038-l1: Enable parent IRQ if necessary
clk: clk-gpio: propagate rate change to parent
clk: qcom: Allow constant ratio freq tables for rcg
clk: qcom: smd: Add missing pnoc clock
f2fs: fix to update dir's i_pino during cross_rename
scsi: lpfc: Fix duplicate unreg_rpi error in port offline flow
scsi: lpfc: Fix unexpected error messages during RSCN handling
scsi: tracing: Fix handling of TRANSFER LENGTH == 0 for READ(6) and WRITE(6)
jbd2: Fix statistics for the number of logged blocks
ext4: iomap that extends beyond EOF should be marked dirty
ext4: update direct I/O read lock pattern for IOCB_NOWAIT
powerpc/book3s64/hash: Add cond_resched to avoid soft lockup warning
powerpc/security/book3s64: Report L1TF status in sysfs
selftests/powerpc: Skip tm-signal-sigreturn-nt if TM not available
dtc: Use pkg-config to locate libyaml
clocksource/drivers/timer-of: Use unique device name instead of timer
clocksource/drivers/asm9260: Add a check for of_clk_get
leds: trigger: netdev: fix handling on interface rename
leds: an30259a: add a check for devm_regmap_init_i2c
leds: lm3692x: Handle failure to probe the regulator
dmaengine: fsl-qdma: Handle invalid qdma-queue0 IRQ
dma-mapping: fix handling of dma-ranges for reserved memory (again)
dma-mapping: Add vmap checks to dma_map_single()
dma-debug: add a schedule point in debug_dma_dump_mappings()
powerpc/tools: Don't quote $objdump in scripts
selftests/powerpc: Fixup clobbers for TM tests
Input: st1232 - do not reset the chip too early
powerpc/pseries: Don't fail hash page table insert for bolted mapping
powerpc/pseries: Mark accumulate_stolen_time() as notrace
scsi: hisi_sas: Delete the debugfs folder of hisi_sas when the probe fails
scsi: hisi_sas: Replace in_softirq() check in hisi_sas_task_exec()
scsi: csiostor: Don't enable IRQs too early
scsi: lpfc: Fix SLI3 hba in loop mode not discovering devices
scsi: lpfc: Fix hardlockup in lpfc_abort_handler
scsi: target: compare full CHAP_A Algorithm strings
dmaengine: xilinx_dma: Clear desc_pendingcount in xilinx_dma_reset
iommu/tegra-smmu: Fix page tables in > 4 GiB memory
iommu: rockchip: Free domain on .domain_free
platform/x86: peaq-wmi: switch to using polled mode of input devices
tools/power/x86/intel-speed-select: Remove warning for unused result
powerpc/papr_scm: Fix an off-by-one check in papr_scm_meta_{get, set}
f2fs: fix to update time in lazytime mode
Input: atmel_mxt_ts - disable IRQ across suspend
scsi: lpfc: Fix list corruption in lpfc_sli_get_iocbq
gpio: mxc: Only get the second IRQ when there is more than one IRQ
scsi: mpt3sas: Reject NVMe Encap cmnds to unsupported HBA
scsi: lpfc: Fix locking on mailbox command completion
scsi: mpt3sas: Fix clear pending bit in ioctl status
scsi: lpfc: Fix discovery failures when target device connectivity bounces
scsi: lpfc: Fix spinlock_irq issues in lpfc_els_flush_cmd()
Revert "MIPS: futex: Emit Loongson3 sync workarounds within asm"
Revert "MIPS: futex: Restore \n after sync instructions"
UPSTREAM: exit: panic before exit_mm() on global init exit
ANDROID: serdev: Fix platform device support
Conflicts:
Documentation/devicetree/bindings
Documentation/devicetree/bindings/Makefile
kernel/time/hrtimer.c
Change-Id: I271162549a080e2b747572f5c87cfd8fa111da51
Signed-off-by: Blagovest Kolenichev <bkolenichev@codeaurora.org>
Update the ABI snapshot due to clean the ABI surface.
Change-Id: I4970f36960d95715fffaf45b1a5eea7d92ed3f44
Signed-off-by: Raghavendra Rao Ananta <rananta@codeaurora.org>
Add debugfs for pmic_glink device to filter out logs based on
filter and enable parameters. This is useful for debugging.
E.g. to filter out logs for PD Altmode driver whose message
owner id is 32780,
echo 1 > /sys/kernel/debug/soc:qcom,pmic_glink/enable
echo 32780 > /sys/kernel/debug/soc:qcom,pmic_glink/filter
E.g. to print out logs for all message owners,
echo 1 > /sys/kernel/debug/soc:qcom,pmic_glink/enable
echo 65535 > /sys/kernel/debug/soc:qcom,pmic_glink/filter
Please note that enabling these logs would print logs on both
Tx and Rx path from the clients of PMIC Glink device.
Change-Id: I63c1c90657b2044fe755cfd8fda73d5f942cad76
Signed-off-by: Subbaraman Narayanamurthy <subbaram@codeaurora.org>
If needed, the IOMMU device can define a CX GDSC to enable before
turning on clocks to directly write IOMMU registers.
Change-Id: Ic0dedbadc8479cc36d5603d6d8a2c93133fda668
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Move ownership of the IOMMU clocks to the KGSL device and get them from
there during IOMMU probe. This ensures that the clocks are owned by a
proper platform device that can sync_state() after we are done probing.
Change-Id: Ic0dedbad97783984f7d8d1c3c2252a1a8a812d77
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Traditionally the GMU/RGMU device has been a child of the adreno device
and probed manually. That has worked despite the fact that the GMU/RGMU
device has resources of its own and it really should be a top level
platform device.
Finally do the right thing, make GMU a top level platform device and
driver and bind it to Adreno as a component. This will make sure we
get the advantages of all of the platform level initialization without
having to code in half a dozen hacks to make everything work and then
break again next kernel version.
Change-Id: Ic0dedbadf6776e1a4f5e0eb1e093241e5103b3ff
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
The KGSL bus probe was not quite resilient for a deferred probe. Move
a memory allocation to devm and make sure that the interconnect handle
is properly put away on failure.
Change-Id: Ic0dedbad261c483a157161a43da0b54459ed46ff
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
This reverts commit 603d76db80.
This change is already present in the snapshot. This revert
will remove duplicate code.
Change-Id: Id4588031a93c041a9480ac8ef33b6d177a4a2477
Signed-off-by: Tony Truong <truong@codeaurora.org>
Enable CPU isolation cooling device driver. This driver will enable the
thermal framework to isolate the CPU cores on thermal conditions.
Change-Id: Ic1f04b7d341c6165d5e971f7dc03ad27bc721897
Signed-off-by: Ram Chandrasekar <rkumbako@codeaurora.org>
Add driver to toggle NFC GPIOs for I3C enumeration to
pass for NFC controller.
Change-Id: I3eef5b8a327c2c70355c03574b06bf67e26b4419
Signed-off-by: Karthik Poosa <kpoosa2@codeaurora.org>
The minidump driver's Kconfig dependency was incorrectly set
to POWER_RESET_QCOM, instead of POWER_RESET_MSM. The build
system sees this as an unknown Kconfig option and as a result,
the minidump driver never gets a chance to be enabled. Hence,
reassign the dependency to the correct Kconfig.
Change-Id: I7ef9f059954b1b6fe2972257435a4eee3bd8ef74
Signed-off-by: Raghavendra Rao Ananta <rananta@codeaurora.org>
Minidump is not yet appropriate for GKI due to missing exported symbols.
Enable statically in Lahaina QGKI for now.
Change-Id: I6bd0db37a8e1c3281249891a432cd0b4b75fe991
Signed-off-by: Elliot Berman <eberman@codeaurora.org>
Enable qbt_handler as a DLKM in the GKI defconfig.
Change-Id: Ib992e9f764ee021518ca7a5533c983d566319129
Signed-off-by: Abir Ghosh <abirg@codeaurora.org>
Add changes to make qbt_handler as DLKM for GKI
Change-Id: I0f4908c01715847aebfc2e440f90b84a16410b1e
Signed-off-by: Rakesh Pallerla <rakeshp@qti.qualcomm.com>
Signed-off-by: Abir Ghosh <abirg@codeaurora.org>
Exclude 4 bytes which holds the size of the buffer while calculating
the actual buffer size to avoid OOB write.
Change-Id: I79704181e0bafc905064621d7cd6e682cca15982
Signed-off-by: suchawla <suchawla@codeaurora.org>
Signed-off-by: Govindaraj Rajagopal <grajagop@codeaurora.org>
Update the frequencies of the clocks from Hz to KHz so that they
match the frequencies requested by the DisplayPort driver. This
ensures that the clock framework correctly scales parent clock
values for all DisplayPort clock requests.
Change-Id: I46be984340902f1facba59ad966e0f2ad1b1b006
Signed-off-by: Tatenda Chipeperekwa <tatendac@codeaurora.org>
Crypto fix to match v5-v6 update which will allow
compilation outside SCSI_UFS_CRYPTO config.
Change-Id: Ic722f17c077c2369303c271a0177d665fe6dcae7
Signed-off-by: Gaurav Kashyap <gaurkash@codeaurora.org>
qsmmuv500_iova_to_phys_hard() takes in the IOMMU domain as an
input parameter, but only uses it to derive the SMMU domain.
In preparation for moving this function out to a file where
to_smmu_domain() will not be available, require that the input
argument be an SMMU domain, instead of the IOMMU domain. This
change has no functional impact.
Change-Id: I2ac7efbae70df2378dc382d697dcfe7a676b85f2
Signed-off-by: Isaac J. Manjarres <isaacm@codeaurora.org>
DOMAIN_ATTR_DEBUG is queried by the qsmmuv500_iova_to_phys()
function, which will move outside of arm-smmu.c, requiring
arm_smmu_domain_get_attr() to be exposed, which doesn't make
sense for only one attribute.
Thus, remove this domain attribute and inform the SMMU driver
that it is operating on a debug domain by introducing a new
marker in the msm_iommu_domain for debug domains. This patch
has no functional impact.
Change-Id: I86dfbadab0a74f9018acc602eea28b2a3a2c38b5
Signed-off-by: Isaac J. Manjarres <isaacm@codeaurora.org>
arm_smmu_init() has retained some workarounds for
back when it could be invoked multiple times as either
part of the initcall sequences, or through the IOMMU
OF table. This was necessary before the driver
core handled dependencies between an IOMMU master
and an IOMMU device. Since this is no longer the case,
we can remove this logic. Also, while we're here,
unregister the TBU driver if the SMMU driver cannot
be registered.
Change-Id: I5c2ed4a046c63babb3bd60390c468a4e5e27de6e
Signed-off-by: Isaac J. Manjarres <isaacm@codeaurora.org>
The current TLB sync timeout handling doesn't make sense
in the context of non QSMMU-V500 SMMU V2 implementations, so
introduce the notion of architecture specific handling
for TLB sync timeouts.
Change-Id: Ibceb5422d6488f9e67cc4c4ed968348ad92aa834
Signed-off-by: Isaac J. Manjarres <isaacm@codeaurora.org>
