[5.9,022/391] ionic: no rx flush in deinit

Message ID	20201103203349.379559871@linuxfoundation.org
State	New
Headers	show Return-Path: <SRS0=OO/H=EJ=vger.kernel.org=stable-owner@kernel.org> From: Greg Kroah-Hartman <gregkh@linuxfoundation.org> To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>, stable@vger.kernel.org, Shannon Nelson <snelson@pensando.io>, Jakub Kicinski <kuba@kernel.org>, Sasha Levin <sashal@kernel.org> Subject: [PATCH 5.9 022/391] ionic: no rx flush in deinit Date: Tue, 3 Nov 2020 21:31:13 +0100 Message-Id: <20201103203349.379559871@linuxfoundation.org> In-Reply-To: <20201103203348.153465465@linuxfoundation.org> References: <20201103203348.153465465@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk
Series	None \| expand [5.9,002/391] xen/events: add a proper barrier to 2-level uevent unmasking [5.9,004/391] xen/events: add a new "late EOI" evtchn framework [5.9,005/391] xen/blkback: use lateeoi irq binding [5.9,008/391] xen/pvcallsback: use lateeoi irq binding [5.9,010/391] xen/events: switch user event channels to lateeoi model [5.9,014/391] firmware: arm_scmi: Fix ARCH_COLD_RESET [5.9,015/391] firmware: arm_scmi: Expand SMC/HVC message pool to more than one [5.9,016/391] tee: client UUID: Skip REE kernel login method as well [5.9,017/391] firmware: arm_scmi: Add missing Rx size re-initialisation [5.9,018/391] x86/unwind/orc: Fix inactive tasks with stack pointer in %sp on GCC 10 compiled ker... [5.9,022/391] ionic: no rx flush in deinit [5.9,023/391] RDMA/mlx5: Fix devlink deadlock on net namespace deletion [5.9,027/391] afs: Fix afs_launder_page to not clear PG_writeback [5.9,029/391] ata: sata_nv: Fix retrieving of active qcs [5.9,031/391] afs: Fix to take ref on page when PG_private is set [5.9,032/391] afs: Fix page leak on afs_write_begin() failure [5.9,035/391] afs: Alter dirty range encoding in page->private [5.9,036/391] afs: Fix afs_invalidatepage to adjust the dirty region [5.9,039/391] interconnect: qcom: sdm845: Enable keepalive for the MM1 BCM [5.9,042/391] futex: Fix incorrect should_fail_futex() handling [5.9,043/391] powerpc/vmemmap: Fix memory leak with vmemmap list allocation failures. [5.9,045/391] RDMA/core: Change how failing destroy is handled during uobj abort [5.9,047/391] powerpc/watchpoint/ptrace: Fix SETHWDEBUG when CONFIG_HAVE_HW_BREAKPOINT=N [5.9,050/391] sparc64: remove mm_cpumask clearing to fix kthread_use_mm race [5.9,051/391] f2fs: add trace exit in exception path [5.9,052/391] f2fs: do sanity check on zoned block device path [5.9,053/391] f2fs: fix uninit-value in f2fs_lookup [5.9,055/391] s390/startup: avoid save_area_sync overflow [5.9,056/391] f2fs: compress: fix to disallow enabling compress on non-empty file [5.9,057/391] s390/ap/zcrypt: revisit ap and zcrypt error handling [5.9,058/391] um: change sigio_spinlock to a mutex [5.9,061/391] powerpc/64s: handle ISA v3.1 local copy-paste context switches [5.9,063/391] NFS4: Fix oops when copy_file_range is attempted with NFS4.0 source [5.9,068/391] xfs: change the order in which child and parent defer ops are finished [5.9,070/391] io_uring: dont set COMP_LOCKED if wont put [5.9,071/391] ath10k: fix retry packets update in station dump [5.9,073/391] drm/ast: Separate DRM driver from PCI code [5.9,076/391] ath10k: start recovery process when payload length exceeds max htc length for sdio [5.9,078/391] drm/scheduler: Scheduler priority fixes (v2) [5.9,081/391] selftests/x86/fsgsbase: Reap a forgotten child [5.9,082/391] drm/bridge_connector: Set default status connected for eDP connectors [5.9,084/391] ASoC: AMD: Clean kernel log from deferred probe error messages [5.9,085/391] misc: fastrpc: fix common struct sg_table related issues [5.9,086/391] staging: wfx: fix potential use before init [5.9,089/391] media: tw5864: check status of tw5864_frameinterval_get [5.9,092/391] mmc: via-sdmmc: Fix data race bug [5.9,093/391] drm/bridge/synopsys: dsi: add support for non-continuous HS clock [5.9,094/391] brcmfmac: increase F2 watermark for BCM4329 [5.9,095/391] arm64: topology: Stop using MPIDR for topology information [5.9,098/391] selftests/powerpc: Make using_hash_mmu() work on Cell & PowerMac [5.9,099/391] kgdb: Make "kgdbcon" work properly with "kgdb_earlycon" [5.9,101/391] drm: exynos: fix common struct sg_table related issues [5.9,102/391] xen: gntdev: fix common struct sg_table related issues [5.9,103/391] drm: lima: fix common struct sg_table related issues [5.9,104/391] drm: panfrost: fix common struct sg_table related issues [5.9,105/391] media: uvcvideo: Fix dereference of out-of-bound list iterator [5.9,107/391] selftests/bpf: Define string const as global for test_sysctl_prog.c [5.9,109/391] samples/bpf: Fix possible deadlock in xdpsock [5.9,113/391] USB: adutux: fix debugging [5.9,115/391] coresight: Make sysfs functional on topologies with per core sink [5.9,116/391] drm/amdgpu: No sysfs, not an error condition [5.9,119/391] SUNRPC: Mitigate cond_resched() in xprt_transmit() [5.9,123/391] habanalabs: remove security from ARB_MST_QUIET register [5.9,124/391] xfs: dont free rt blocks when were doing a REMAP bunmapi call [5.9,126/391] ACPI: Add out of bounds and numa_off protections to pxm_to_node() [5.9,128/391] brcmfmac: Fix warning message after dongle setup failed [5.9,129/391] ath11k: Use GFP_ATOMIC instead of GFP_KERNEL in ath11k_dp_htt_get_ppdu_desc [5.9,131/391] ath11k: change to disable softirqs for ath11k_regd_update to solve deadlock [5.9,132/391] drivers/net/wan/hdlc_fr: Correctly handle special skb->protocol values [5.9,135/391] bus/fsl_mc: Do not rely on caller to provide non NULL mc_io [5.9,136/391] ACPI: HMAT: Fix handling of changes from ACPI 6.2 to ACPI 6.3 [5.9,139/391] drm/amd/display: Avoid set zero in the requested clk [5.9,140/391] ARC: [dts] fix the errors detected by dtbs_check [5.9,142/391] btrfs: fix replace of seed device [5.9,144/391] f2fs: fix to set SBI_NEED_FSCK flag for inconsistent inode [5.9,146/391] rpmsg: glink: Use complete_all for open states [5.9,149/391] nfsd: rename delegation related tracepoints to make them less confusing [5.9,152/391] ceph: encode inodes parent/d_name in cap reconnect message [5.9,154/391] jbd2: avoid transaction reuse after reformatting [5.9,156/391] KVM: PPC: Book3S HV: Do not allocate HPT for a nested guest [5.9,160/391] gfs2: use-after-free in sysfs deregistration [5.9,161/391] gfs2: add validation checks for size of superblock [5.9,163/391] cifs: handle -EINTR in cifs_setattr [5.9,166/391] memory: emif: Remove bogus debugfs error handling [5.9,167/391] ARM: dts: s5pv210: Enable audio on Aries boards [5.9,169/391] ARM: dts: s5pv210: move fixed clocks under root node [5.9,171/391] ARM: dts: s5pv210: remove dedicated audio-subsystem node [5.9,173/391] ARM: dts: s5pv210: align SPI GPIO node name with dtschema in Aries [5.9,174/391] soc: qcom: rpmh-rsc: Sleep waiting for tcs slots to be free [5.9,175/391] soc: ti: k3: ringacc: add am65x sr2.0 support [5.9,176/391] bindings: soc: ti: soc: ringacc: remove ti,dma-ring-reset-quirk [5.9,177/391] firmware: arm_scmi: Move scmi bus init and exit calls into the driver [5.9,178/391] arm64: dts: qcom: kitakami: Temporarily disable SDHCI1 [5.9,179/391] nbd: make the config put is called before the notifying the waiter [5.9,182/391] vmlinux.lds.h: Add PGO and AutoFDO input sections [5.9,183/391] irqchip/loongson-htvec: Fix initial interrupt clearing [5.9,185/391] md/raid5: fix oops during stripe resizing [5.9,189/391] seccomp: Make duplicate listener detection non-racy [5.9,191/391] perf/x86/intel: Fix Ice Lake event constraint table [5.9,192/391] perf/x86/amd: Fix sampling Large Increment per Cycle events [5.9,194/391] perf/x86/amd/ibs: Dont include randomized bits in get_ibs_op_count() [5.9,195/391] perf/x86/amd/ibs: Fix raw sample data accumulation [5.9,197/391] spi: sprd: Release DMA channel also on probe deferral [5.9,198/391] extcon: ptn5150: Fix usage of atomic GPIO with sleeping GPIO chips [5.9,200/391] hwmon: (pmbus/max34440) Fix OC fault limits [5.9,203/391] ACPI: configfs: Add missing config_item_put() to fix refcount leak [5.9,204/391] NFS: fix nfs_path in case of a rename retry [5.9,206/391] ACPI / extlog: Check for RDMSR failure [5.9,208/391] ACPI: debug: dont allow debugging when ACPI is disabled [5.9,210/391] ACPI: EC: PM: Flush EC work unconditionally after wakeup [5.9,212/391] acpi-cpufreq: Honor _PSD table setting on new AMD CPUs [5.9,213/391] io-wq: assign NUMA node locality if appropriate [5.9,215/391] fs/kernel_read_file: Remove FIRMWARE_PREALLOC_BUFFER enum [5.9,216/391] scsi: mptfusion: Fix null pointer dereferences in mptscsih_remove() [5.9,218/391] scsi: qla2xxx: Fix reset of MPI firmware [5.9,222/391] btrfs: improve device scanning messages [5.9,223/391] btrfs: qgroup: fix qgroup meta rsv leak for subvolume operations [5.9,225/391] btrfs: tracepoints: output proper root owner for trace_find_free_extent() [5.9,226/391] btrfs: reschedule if necessary when logging directory items [5.9,228/391] btrfs: send, recompute reference path after orphanization of a directory [5.9,229/391] btrfs: use kvzalloc() to allocate clone_roots in btrfs_ioctl_send() [5.9,231/391] btrfs: reschedule when cloning lots of extents [5.9,233/391] btrfs: skip devices without magic signature when mounting [5.9,234/391] btrfs: tree-checker: validate number of chunk stripes and parity [5.9,236/391] btrfs: fix readahead hang and use-after-free after removing a device [5.9,237/391] btrfs: drop the path before adding block group sysfs files [5.9,240/391] usb: dwc3: ep0: Fix ZLP for OUT ep0 requests [5.9,242/391] usb: dwc3: gadget: Reclaim extra TRBs after request completion [5.9,243/391] usb: dwc3: core: add phy cleanup for probe error handling [5.9,246/391] usb: dwc3: gadget: END_TRANSFER before CLEAR_STALL command [5.9,248/391] usb: cdc-acm: fix cooldown mechanism [5.9,250/391] usb: host: fsl-mph-dr-of: check return of dma_set_mask() [5.9,252/391] USB: apple-mfi-fastcharge: dont probe unhandled devices [5.9,255/391] vt: keyboard, extend func_buf_lock to readers [5.9,257/391] HID: wacom: Avoid entering wacom_wac_pen_report for pad / battery [5.9,258/391] x86/mce: Allow for copy_mc_fragile symbol checksum to be generated [5.9,263/391] powerpc: Fix random segfault when freeing hugetlb range [5.9,265/391] dmaengine: dma-jz4780: Fix race in jz4780_dma_tx_status [5.9,267/391] drm/shme-helpers: Fix dma_buf_mmap forwarding bug [5.9,269/391] iio: adc: at91-sama5d2_adc: fix DMA conversion crash [5.9,270/391] iio:imu:inv_mpu6050 Fix dma and ts alignment and data leak issues. [5.9,272/391] iio:light:si1145: Fix timestamp alignment and prevent data leak. [5.9,273/391] iio: adc: gyroadc: fix leak of device node iterator [5.9,275/391] iio:adc:ti-adc0832 Fix alignment issue with timestamp [5.9,276/391] iio:adc:ti-adc12138 Fix alignment issue with timestamp [5.9,277/391] iio:imu:st_lsm6dsx Fix alignment and data leak issues [5.9,278/391] iio:gyro:itg3200: Fix timestamp alignment and prevent data leak. [5.9,281/391] rcu-tasks: Fix low-probability task_struct leak [5.9,282/391] rcu-tasks: Enclose task-list scan in rcu_read_lock() [5.9,283/391] MIPS: DEC: Restore bootmem reservation for firmware working memory area [5.9,286/391] powerpc/rtas: Restrict RTAS requests from userspace [5.9,287/391] powerpc: Warn about use of smt_snooze_delay [5.9,288/391] powerpc/memhotplug: Make lmb size 64bit [5.9,292/391] powerpc: Fix undetected data corruption with P9N DD2.1 VSX CI load emulation [5.9,296/391] io_uring: use type appropriate io_kiocb handler for double poll [5.9,298/391] gfs2: Make sure we dont miss any delayed withdraws [5.9,299/391] gfs2: Only access gl_delete for iopen glocks [5.9,301/391] NFSv4.2: support EXCHGID4_FLAG_SUPP_FENCE_OPS 4.2 EXCHANGE_ID flag [5.9,302/391] NFSD: Add missing NFSv2 .pc_func methods [5.9,304/391] ubifs: xattr: Fix some potential memory leaks while iterating entries [5.9,305/391] ubifs: journal: Make sure to not dirty twice for auth nodes [5.9,307/391] ubifs: Dont parse authentication mount options in remount process [5.9,308/391] ubifs: mount_ubifs: Release authentication resource in error handling path [5.9,312/391] ubi: check kthread_should_stop() after the setting of task state [5.9,313/391] ia64: fix build error with !COREDUMP [5.9,315/391] i2c: imx: Fix external abort on interrupt in exit paths [5.9,318/391] drm/amd/display: Increase timeout for DP Disable [5.9,319/391] drm/amdgpu: vcn and jpeg ring synchronization [5.9,320/391] drm/amdgpu: update golden setting for sienna_cichlid [5.9,322/391] drm/amdkfd: Use same SQ prefetch setting as amdgpu [5.9,325/391] drm/amdgpu: increase the reserved VM size to 2MB [5.9,326/391] drm/amd/display: Dont invoke kgdb_breakpoint() unconditionally [5.9,328/391] ceph: promote to unsigned long long before shifting [5.9,332/391] PCI: qcom: Make sure PCIe is reset before init for rev 2.1.0 [5.9,334/391] intel_idle: Ignore _CST if control cannot be taken from the platform [5.9,337/391] cpufreq: Introduce CPUFREQ_NEED_UPDATE_LIMITS driver flag [5.9,338/391] cpufreq: intel_pstate: Avoid missing HWP max updates in passive mode [5.9,340/391] ext4: fix leaking sysfs kobject after failed mount [5.9,343/391] ext4: fix invalid inode checksum [5.9,344/391] ext4: fix superblock checksum calculation race [5.9,346/391] ext4: fix bdev write error check failed when mount fs with ro [5.9,347/391] ext4: fix bs < ps issue reported with dioread_nolock mount opt [5.9,349/391] drm/ttm: fix eviction valuable range check. [5.9,350/391] mmc: sdhci-of-esdhc: make sure delay chain locked for HS400 [5.9,352/391] mmc: sdhci: Use Auto CMD Auto Select only when v4_mode is true [5.9,356/391] drm/amdgpu/swsmu: drop smu i2c bus on navi1x [5.9,358/391] drm/amd/pm: fix pp_dpm_fclk [5.9,359/391] drm/amd/swsmu: add missing feature map for sienna_cichlid [5.9,363/391] arm64: berlin: Select DW_APB_TIMER_OF [5.9,365/391] hil/parisc: Disable HIL driver when it gets stuck [5.9,366/391] arm: dts: mt7623: add missing pause for switchport [5.9,367/391] ARM: aspeed: g5: Do not set sirq polarity [5.9,369/391] ARM: config: aspeed: Fix selection of media drivers [5.9,371/391] ARM: s3c24xx: fix missing system reset [5.9,372/391] arm64: Change .weak to SYM_FUNC_START_WEAK_PI for arch/arm64/lib/mem*.S [5.9,373/391] arm64: dts: marvell: espressobin: Add ethernet switch aliases [5.9,375/391] coresight: cti: Initialize dynamic sysfs attributes [5.9,377/391] device property: Dont clear secondary pointer for shared primary firmware node [5.9,379/391] KVM: arm64: Fix AArch32 handling of DBGD{CCINT,SCRext} and DBGVCR [5.9,383/391] staging: octeon: repair "fixed-link" support [5.9,384/391] staging: octeon: Drop on uncorrectable alignment or FCS error [5.9,386/391] cpufreq: schedutil: Always call driver if CPUFREQ_NEED_UPDATE_LIMITS is set [5.9,388/391] vdpa/mlx5: Fix error return in map_direct_mr() [5.9,389/391] time: Prevent undefined behaviour in timespec64_to_ns() [5.9,390/391] time/sched_clock: Mark sched_clock_read_begin/retry() as notrace

Message ID

20201103203349.379559871@linuxfoundation.org

State

New

Headers

From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Shannon Nelson <snelson@pensando.io>,
	Jakub Kicinski <kuba@kernel.org>, Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.9 022/391] ionic: no rx flush in deinit
Date: Tue,  3 Nov 2020 21:31:13 +0100
Message-Id: <20201103203349.379559871@linuxfoundation.org>
In-Reply-To: <20201103203348.153465465@linuxfoundation.org>
References: <20201103203348.153465465@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

Series

None | expand

Commit Message

Greg Kroah-Hartman Nov. 3, 2020, 8:31 p.m. UTC

[ Upstream commit 43ecf7b46f2688fd37909801aee264f288b3917b ]

Kmemleak pointed out to us that ionic_rx_flush() is sending
skbs into napi_gro_XXX with a disabled napi context, and these
end up getting lost and leaked.  We can safely remove the flush.

Fixes: 0f3154e6bcb3 ("ionic: Add Tx and Rx handling")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/net/ethernet/pensando/ionic/ionic_lif.c  |  1 -
 drivers/net/ethernet/pensando/ionic/ionic_txrx.c | 13 -------------
 drivers/net/ethernet/pensando/ionic/ionic_txrx.h |  1 -
 3 files changed, 15 deletions(-)

diff --git a/drivers/net/ethernet/pensando/ionic/ionic_lif.c b/drivers/net/ethernet/pensando/ionic/ionic_lif.c
index 26988ad7ec979..8867d4ac871c1 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_lif.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_lif.c
@@ -1512,7 +1512,6 @@  static void ionic_txrx_deinit(struct ionic_lif *lif)
 	if (lif->rxqcqs) {
 		for (i = 0; i < lif->nxqs; i++) {
 			ionic_lif_qcq_deinit(lif, lif->rxqcqs[i].qcq);
-			ionic_rx_flush(&lif->rxqcqs[i].qcq->cq);
 			ionic_rx_empty(&lif->rxqcqs[i].qcq->q);
 		}
 	}
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_txrx.c b/drivers/net/ethernet/pensando/ionic/ionic_txrx.c
index def65fee27b5a..39e85870c15e9 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_txrx.c
+++ b/drivers/net/ethernet/pensando/ionic/ionic_txrx.c
@@ -253,19 +253,6 @@  static bool ionic_rx_service(struct ionic_cq *cq, struct ionic_cq_info *cq_info)
 	return true;
 }
 
-void ionic_rx_flush(struct ionic_cq *cq)
-{
-	struct ionic_dev *idev = &cq->lif->ionic->idev;
-	u32 work_done;
-
-	work_done = ionic_cq_service(cq, cq->num_descs,
-				     ionic_rx_service, NULL, NULL);
-
-	if (work_done)
-		ionic_intr_credits(idev->intr_ctrl, cq->bound_intr->index,
-				   work_done, IONIC_INTR_CRED_RESET_COALESCE);
-}
-
 static struct page *ionic_rx_page_alloc(struct ionic_queue *q,
 					dma_addr_t *dma_addr)
 {
diff --git a/drivers/net/ethernet/pensando/ionic/ionic_txrx.h b/drivers/net/ethernet/pensando/ionic/ionic_txrx.h
index a5883be0413f6..7667b72232b8a 100644
--- a/drivers/net/ethernet/pensando/ionic/ionic_txrx.h
+++ b/drivers/net/ethernet/pensando/ionic/ionic_txrx.h
@@ -4,7 +4,6 @@ 
 #ifndef _IONIC_TXRX_H_
 #define _IONIC_TXRX_H_
 
-void ionic_rx_flush(struct ionic_cq *cq);
 void ionic_tx_flush(struct ionic_cq *cq);
 
 void ionic_rx_fill(struct ionic_queue *q);

[5.9,022/391] ionic: no rx flush in deinit

Commit Message

Patch