[5.7,161/376] kgdb: Fix spurious true from in_dbg_master()

Message ID	20200619141717.943538361@linuxfoundation.org
State	New
Headers	show Delivered-To: patch@linaro.org Received-SPF: pass (google.com: domain of stable-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18; From: Greg Kroah-Hartman <gregkh@linuxfoundation.org> To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>, stable@vger.kernel.org, Will Deacon <will@kernel.org>, Douglas Anderson <dianders@chromium.org>, Daniel Thompson <daniel.thompson@linaro.org>, Sasha Levin <sashal@kernel.org> Subject: [PATCH 5.7 161/376] kgdb: Fix spurious true from in_dbg_master() Date: Fri, 19 Jun 2020 16:31:19 +0200 Message-Id: <20200619141717.943538361@linuxfoundation.org> In-Reply-To: <20200619141710.350494719@linuxfoundation.org> References: <20200619141710.350494719@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Sender: stable-owner@vger.kernel.org Precedence: bulk
Series	None \| expand [5.7,003/376] ath10k: Fix the race condition in firmware dump work queue [5.7,006/376] media: imx: imx7-mipi-csis: Cleanup and fix subdev pad format handling [5.7,007/376] crypto: ccp -- dont "select" CONFIG_DMADEVICES [5.7,010/376] media: vicodec: Fix error codes in probe function [5.7,013/376] objtool: Ignore empty alternatives [5.7,014/376] drm/amd/display: Force watermark value propagation [5.7,015/376] drm/amd/display: fix virtual signal dsc setup [5.7,016/376] spi: spi-mem: Fix Dual/Quad modes on Octal-capable devices [5.7,018/376] drm/bridge: panel: Return always an error pointer in drm_panel_bridge_add() [5.7,019/376] net: ethernet: ti: fix return value check in k3_cppi_desc_pool_create_name() [5.7,020/376] arm64/kernel: Fix range on invalidating dcache for boot page tables [5.7,022/376] libbpf: Fix memory leak and possible double-free in hashmap__clear [5.7,026/376] net: atlantic: make hw_get_regs optional [5.7,027/376] net: ena: fix error returning in ena_com_get_hash_function() [5.7,030/376] arm64: cacheflush: Fix KGDB trap detection [5.7,032/376] libperf evlist: Fix a refcount leak [5.7,033/376] rtw88: fix an issue about leak system resources [5.7,034/376] spi: dw: Zero DMA Tx and Rx configurations on stack [5.7,035/376] soc: fsl: dpio: properly compute the consumer index [5.7,038/376] arm64: insn: Fix two bugs in encoding 32-bit logical immediates [5.7,039/376] tools/power/x86/intel-speed-select: Fix CLX-N package information output [5.7,040/376] mt76: mt7615: fix aid configuration in mt7615_mcu_wtbl_generic_tlv [5.7,041/376] block: reset mapping if failed to update hardware queue count [5.7,043/376] lockdown: Allow unprivileged users to see lockdown status [5.7,044/376] ixgbe: Fix XDP redirect on archs with PAGE_SIZE above 4K [5.7,045/376] platform/x86: dell-laptop: dont register micmute LED if there is no token [5.7,046/376] cpuidle: psci: Fixup execution order when entering a domain idle state [5.7,048/376] io_uring: cleanup io_poll_remove_one() logic [5.7,049/376] media: i2c: imx219: Fix a bug in imx219_enum_frame_size [5.7,051/376] kgdb: Disable WARN_CONSOLE_UNLOCKED for all kgdb [5.7,052/376] kgdb: Prevent infinite recursive entries to the debugger [5.7,053/376] pmu/smmuv3: Clear IRQ affinity hint on device removal [5.7,054/376] ath11k: Fix some resource leaks in error path in ath11k_thermal_register() [5.7,058/376] spi: dw: Enable interrupts in accordance with DMA xfer mode [5.7,059/376] clocksource/drivers/timer-versatile: Clear OF_POPULATED flag [5.7,061/376] clocksource: dw_apb_timer_of: Fix missing clockevent timers [5.7,065/376] ARM: 8978/1: mm: make act_mm() respect THREAD_SIZE [5.7,066/376] batman-adv: Revert "disable ethtool link speed detection when auto negotiation off" [5.7,069/376] ice: Fix for memory leaks and modify ICE_FREE_CQ_BUFS [5.7,072/376] mmc: meson-mx-sdio: trigger a soft reset after a timeout or CRC error [5.7,074/376] Bluetooth: hci_qca: Fix suspend/resume functionality failure [5.7,075/376] spi: dw: Fix Rx-only DMA transfers [5.7,077/376] x86/kvm/hyper-v: Explicitly align hcall param for kvm_hyperv_exit [5.7,080/376] x86: fix vmap arguments in map_irq_stack [5.7,082/376] ubsan: entirely disable alignment checks under UBSAN_TRAP [5.7,083/376] ath11k: fix error message to correctly report the command that failed [5.7,085/376] ath11k: Avoid mgmt tx count underflow [5.7,087/376] drm/amd/display: Revert to old formula in set_vtg_params [5.7,088/376] media: staging/intel-ipu3: Implement lock for stream on/off operations [5.7,090/376] spi: Respect DataBitLength field of SpiSerialBusV2() ACPI resource [5.7,091/376] brcmfmac: fix wrong location to get firmware feature [5.7,092/376] regulator: qcom-rpmh: Fix typos in pm8150 and pm8150l [5.7,093/376] tools api fs: Make xxx__mountpoint() more scalable [5.7,095/376] net: mscc: ocelot: deal with problematic MAC_ETYPE VCAP IS2 rules [5.7,099/376] audit: fix a net reference leak in audit_send_reply() [5.7,100/376] media: dvb: return -EREMOTEIO on i2c transfer failure. [5.7,103/376] media: platform: fcp: Set appropriate DMA parameters [5.7,106/376] Bluetooth: btbcm: Add 2 missing models to subver tables [5.7,107/376] audit: fix a net reference leak in audit_list_rules_send() [5.7,108/376] drm/amd/display: Correct updating logic of dcn21s pipe VM flags [5.7,110/376] Drivers: hv: vmbus: Always handle the VMBus messages on CPU0 [5.7,111/376] efi/libstub/random: Align allocate size to EFI_ALLOC_ALIGN [5.7,113/376] bcache: remove a duplicate ->make_request_fn assignment [5.7,117/376] ath11k: use GFP_ATOMIC under spin lock [5.7,118/376] Bluetooth: Adding driver and quirk defs for multi-role LE [5.7,121/376] selftests/bpf: Ensure test flavors use correct skeletons [5.7,123/376] selftests/bpf: Fix memory leak in extract_build_id() [5.7,126/376] selftests/bpf: Fix bpf_link leak in ns_current_pid_tgid selftest [5.7,127/376] selftests/bpf: Add runqslower binary to .gitignore [5.7,129/376] ARM: 8969/1: decompressor: simplify libfdt builds [5.7,130/376] drm/bridge: fix stack usage warning on old gcc [5.7,132/376] net: bcmgenet: Fix WoL with password after deep sleep [5.7,135/376] exit: Move preemption fixup up, move blocking operations down [5.7,136/376] sched/core: Fix illegal RCU from offline CPUs [5.7,138/376] arm64: kexec_file: print appropriate variable [5.7,140/376] iocost_monitor: drop string wrap around numbers when outputting json [5.7,141/376] octeontx2-pf: Fix error return code in otx2_probe() [5.7,143/376] net: lpc-enet: fix error return code in lpc_mii_init() [5.7,144/376] selinux: fix error return code in policydb_read() [5.7,146/376] cpufreq: qcom: fix wrong compatible binding [5.7,147/376] ath10k: fix possible memory leak in ath10k_bmi_lz_data_large() [5.7,148/376] ath11k: fix error return code in ath11k_dp_alloc() [5.7,149/376] media: sun8i: Fix an error handling path in deinterlace_runtime_resume() [5.7,153/376] ath11k: fix kernel panic by freeing the msdu received with invalid length [5.7,154/376] libertas_tf: avoid a null dereference in pointer priv [5.7,155/376] xfs: clean up the error handling in xfs_swap_extents [5.7,157/376] Crypto/chcr: fix for ccm(aes) failed test [5.7,158/376] dsa: sja1105: dynamically allocate stats structure [5.7,159/376] MIPS: Truncate link address into 32bit for 32bit kernel [5.7,161/376] kgdb: Fix spurious true from in_dbg_master() [5.7,163/376] xfs: fix duplicate verification from xfs_qm_dqflush() [5.7,165/376] platform/x86: intel-vbtn: Split keymap into buttons and switches parts [5.7,168/376] iwlwifi: avoid debug max amsdu config overwriting itself [5.7,169/376] nvme: refine the Qemu Identify CNS quirk [5.7,170/376] nvme-fc: avoid gcc-10 zero-length-bounds warning [5.7,172/376] nvme-tcp: use bh_lock in data_ready [5.7,175/376] wcn36xx: Fix error handling path in wcn36xx_probe() [5.7,184/376] mt76: avoid rx reorder buffer overflow [5.7,186/376] brcmfmac: fix WPA/WPA2-PSK 4-way handshake offload and SAE offload failures [5.7,188/376] raid5: remove gfp flags from scribble_alloc() [5.7,189/376] iocost: dont let vrate run wild while theres no saturation signal [5.7,191/376] crypto: blake2b - Fix clang optimization for ARMv7-M [5.7,193/376] net/mlx5e: IPoIB, Drop multicast packets that this interface sent [5.7,195/376] net: ipa: do not clear interrupt in gsi_channel_start() [5.7,196/376] rtlwifi: Fix a double free in _rtl_usb_tx_urb_setup() [5.7,197/376] mwifiex: Fix memory corruption in dump_station [5.7,199/376] x86/boot: Correct relocation destination on old linkers [5.7,201/376] sched: Defend cfs and rt bandwidth quota against overflow [5.7,202/376] mips: MAAR: Use more precise address mask [5.7,204/376] ice: Fix resource leak on early exit from function [5.7,206/376] crypto: stm32/crc32 - fix ext4 chksum BUG_ON() [5.7,208/376] crypto: stm32/crc32 - fix multi-instance [5.7,209/376] drm/amd/powerpay: Disable gfxoff when setting manual mode on picasso and raven [5.7,210/376] drm/amdgpu: Sync with VM root BO when switching VM to CPU update mode [5.7,211/376] selftests/bpf: CONFIG_IPV6_SEG6_BPF required for test_seg6_loop.o [5.7,212/376] selftests/bpf: CONFIG_LIRC required for test_lirc_mode2.sh [5.7,214/376] x86/mm: Stop printing BRK addresses [5.7,215/376] MIPS: Fix exception handler memcpy() [5.7,218/376] btrfs: improve global reserve stealing logic [5.7,221/376] net: ethernet: fec: move GPR register offset and bit into DT [5.7,222/376] nvme-pci: make sure write/poll_queues less or equal then cpu count [5.7,223/376] nvmet: fix memory leak when removing namespaces and controllers concurrently [5.7,226/376] MIPS: Fix IRQ tracing when call handle_fpe() and handle_msa_fpe() [5.7,228/376] mmc: mmci: Switch to mmc_regulator_set_vqmmc() [5.7,230/376] staging: greybus: sdio: Respect the cmd->busy_timeout from the mmc core [5.7,231/376] mmc: owl-mmc: Respect the cmd->busy_timeout from the mmc core [5.7,232/376] mmc: via-sdmmc: Respect the cmd->busy_timeout from the mmc core [5.7,234/376] ice: fix potential double free in probe unrolling [5.7,235/376] ixgbe: fix signed-integer-overflow warning [5.7,237/376] mmc: sdhci-esdhc-imx: fix the mask for tuning start point [5.7,238/376] spi: dw: Return any value retrieved from the dma_transfer callback [5.7,241/376] io_uring: fix overflowed reqs cancellation [5.7,245/376] platform/x86: intel-vbtn: Only blacklist SW_TABLET_MODE on the 9 / "Laptop" chasis-... [5.7,247/376] libbpf: Fix perf_buffer__free() API for sparse allocs [5.7,249/376] bpf: Refactor sockmap redirect code so its easy to reuse [5.7,251/376] selftests/bpf, flow_dissector: Close TAP device FD after the test [5.7,254/376] kasan: stop tests being eliminated as dead code with FORTIFY_SOURCE [5.7,255/376] string.h: fix incompatibility between FORTIFY_SOURCE and KASAN [5.7,256/376] btrfs: free alien device after device add [5.7,258/376] btrfs: fix a race between scrub and block group removal/allocation [5.7,261/376] btrfs: reloc: fix reloc root leak and NULL pointer dereference [5.7,264/376] btrfs: fix wrong file range cleanup after an error filling dealloc range [5.7,265/376] btrfs: fix space_info bytes_may_use underflow after nocow buffered write [5.7,266/376] btrfs: fix space_info bytes_may_use underflow during space cache writeout [5.7,268/376] mm: thp: make the THP mapcount atomic against __split_huge_pmd_locked() [5.7,271/376] mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init [5.7,273/376] ima: Fix ima digest hash table key calculation [5.7,275/376] ima: Evaluate error in init_ima() [5.7,276/376] ima: Directly assign the ima_default_policy pointer to ima_rules [5.7,277/376] ima: Call ima_calc_boot_aggregate() in ima_eventdigest_init() [5.7,279/376] evm: Fix possible memory leak in evm_calc_hmac_or_hash() [5.7,283/376] ext4: fix race between ext4_sync_parent() and rename() [5.7,284/376] PCI: Avoid Pericom USB controller OHCI/EHCI PME# defect [5.7,286/376] PCI: Avoid FLR for AMD Starship USB 3.0 [5.7,288/376] serial: 8250_pci: Move Pericom IDs to pci_ids.h [5.7,289/376] x86/amd_nb: Add AMD family 17h model 60h PCI IDs [5.7,291/376] ima: Set again build_ima_appraise variable [5.7,292/376] PCI: Program MPS for RCiEP devices [5.7,297/376] carl9170: remove P2P_GO support [5.7,298/376] clocksource: Remove obsolete ifdef [5.7,300/376] media: cedrus: Program output format during each run [5.7,302/376] Bluetooth: hci_bcm: respect IRQ polarity from DT [5.7,304/376] b43legacy: Fix case where channel status is corrupted [5.7,307/376] media: ov5640: fix use of destroyed mutex [5.7,308/376] usb: musb: mediatek: add reset FADDR to zero in reset interrupt handle [5.7,309/376] clk: mediatek: assign the initial value to clk_init_data of mtk_mux [5.7,312/376] EDAC/amd64: Add AMD family 17h model 60h PCI IDs [5.7,314/376] iommu/vt-d: Allocate domain info for real DMA sub-devices [5.7,315/376] power: vexpress: add suppress_bind_attrs to true [5.7,316/376] power: supply: core: fix HWMON temperature labels [5.7,317/376] power: supply: core: fix memory leak in HWMON error path [5.7,319/376] pinctrl: samsung: Save/restore eint_mask over suspend for EINT_TYPE GPIOs [5.7,321/376] sparc32: fix register window handling in genregs32_[gs]et() [5.7,322/376] sparc64: fix misuses of access_process_vm() in genregs32_[sg]et() [5.7,324/376] dm crypt: avoid truncating the logical block size [5.7,326/376] memory: samsung: exynos5422-dmc: Fix tFAW timings alignment [5.7,328/376] powerpc/fadump: consider reserved ranges while reserving memory [5.7,330/376] kernel/cpu_pm: Fix uninitted local in cpu_pm [5.7,331/376] ARM: tegra: Correct PL310 Auxiliary Control Register initialization [5.7,334/376] ARM: dts: exynos: Fix GPIO polarity for thr GalaxyS3 CM36651 sensors bus [5.7,335/376] ARM: dts: at91: sama5d2_ptc_ek: fix vbus pin [5.7,336/376] ARM: dts: s5pv210: Set keep-power-in-suspend for SDHCI1 on Aries [5.7,338/376] powerpc/32s: Fix another build failure with CONFIG_PPC_KUAP_DEBUG [5.7,340/376] powerpc/kasan: Fix shadow pages allocation failure [5.7,344/376] kbuild: force to build vmlinux if CONFIG_MODVERSION=y [5.7,345/376] virtio-balloon: Disable free page reporting if page poison reporting is not enabled [5.7,346/376] sunrpc: svcauth_gss_register_pseudoflavor must reject duplicate registrations. [5.7,348/376] block: nr_sects_write(): Disable preemption on seqcount write [5.7,350/376] pwm: lpss: Fix get_state runtime-pm reference handling [5.7,353/376] mtd: rawnand: onfi: Fix redundancy detection check [5.7,354/376] mtd: rawnand: brcmnand: fix hamming oob layout [5.7,356/376] mtd: rawnand: sharpsl: Fix the probe error path [5.7,358/376] mtd: rawnand: xway: Fix the probe error path [5.7,360/376] mtd: rawnand: socrates: Fix the probe error path [5.7,363/376] mtd: rawnand: plat_nand: Fix the probe error path [5.7,365/376] mtd: rawnand: mtk: Fix the probe error path [5.7,367/376] w1: omap-hdq: cleanup to add missing newline for some dev_dbg [5.7,370/376] f2fs: dont leak filename in f2fs_try_convert_inline_dir() [5.7,373/376] perf probe: Fix to check blacklist address correctly [5.7,374/376] perf probe: Check address correctness by map instead of _etext [5.7,376/376] perf symbols: Fix kernel maps for kcore and eBPF

Message ID

20200619141717.943538361@linuxfoundation.org

State

New

Headers

Received-SPF: pass (google.com: domain of stable-owner@vger.kernel.org
	designates 23.128.96.18 as permitted sender)
	client-ip=23.128.96.18; 
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Will Deacon <will@kernel.org>,
	Douglas Anderson <dianders@chromium.org>,
	Daniel Thompson <daniel.thompson@linaro.org>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.7 161/376] kgdb: Fix spurious true from in_dbg_master()
Date: Fri, 19 Jun 2020 16:31:19 +0200
Message-Id: <20200619141717.943538361@linuxfoundation.org>
In-Reply-To: <20200619141710.350494719@linuxfoundation.org>
References: <20200619141710.350494719@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Sender: stable-owner@vger.kernel.org
Precedence: bulk

Series

None | expand

Commit Message

Greg KH June 19, 2020, 2:31 p.m. UTC

From: Daniel Thompson <daniel.thompson@linaro.org>


[ Upstream commit 3fec4aecb311995189217e64d725cfe84a568de3 ]

Currently there is a small window where a badly timed migration could
cause in_dbg_master() to spuriously return true. Specifically if we
migrate to a new core after reading the processor id and the previous
core takes a breakpoint then we will evaluate true if we read
kgdb_active before we get the IPI to bring us to halt.

Fix this by checking irqs_disabled() first. Interrupts are always
disabled when we are executing the kgdb trap so this is an acceptable
prerequisite. This also allows us to replace raw_smp_processor_id()
with smp_processor_id() since the short circuit logic will prevent
warnings from PREEMPT_DEBUG.

Fixes: dcc7871128e9 ("kgdb: core changes to support kdb")
Suggested-by: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20200506164223.2875760-1-daniel.thompson@linaro.org
Reviewed-by: Douglas Anderson <dianders@chromium.org>

Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>

Signed-off-by: Sasha Levin <sashal@kernel.org>

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

-- 
2.25.1

diff --git a/include/linux/kgdb.h b/include/linux/kgdb.h
index b072aeb1fd78..4d6fe87fd38f 100644
--- a/include/linux/kgdb.h
+++ b/include/linux/kgdb.h
@@ -323,7 +323,7 @@  extern void gdbstub_exit(int status);
 extern int			kgdb_single_step;
 extern atomic_t			kgdb_active;
 #define in_dbg_master() \
-	(raw_smp_processor_id() == atomic_read(&kgdb_active))
+	(irqs_disabled() && (smp_processor_id() == atomic_read(&kgdb_active)))
 extern bool dbg_is_early;
 extern void __init dbg_late_init(void);
 extern void kgdb_panic(const char *msg);

[5.7,161/376] kgdb: Fix spurious true from in_dbg_master()

Commit Message

Patch