LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Bodong Wang <bodong@mellanox.com>,
	Parav Pandit <parav@mellanox.com>, Vu Pham <vuhuong@mellanox.com>,
	Saeed Mahameed <saeedm@mellanox.com>,
	Sasha Levin <sashal@kernel.org>,
	netdev@vger.kernel.org, linux-rdma@vger.kernel.org
Subject: [PATCH AUTOSEL 5.1 040/375] net/mlx5: E-Switch, Use atomic rep state to serialize state change
Date: Wed, 22 May 2019 15:15:40 -0400	[thread overview]
Message-ID: <20190522192115.22666-40-sashal@kernel.org> (raw)
In-Reply-To: <20190522192115.22666-1-sashal@kernel.org>

From: Bodong Wang <bodong@mellanox.com>

[ Upstream commit 6f4e02193c9a9ea54dd3151cf97489fa787cd0e6 ]

When the state of rep was introduced, it was also designed to prevent
duplicate unloading of the same rep. Considering the following two
flows when an eswitch manager is at switchdev mode with n VF reps loaded.

+--------------------------------------+--------------------------------+
| cpu-0                                | cpu-1                          |
| --------                             | --------                       |
| mlx5_ib_remove                       | mlx5_eswitch_disable_sriov     |
|  mlx5_ib_unregister_vport_reps       |  esw_offloads_cleanup          |
|   mlx5_eswitch_unregister_vport_reps |   esw_offloads_unload_all_reps |
|    __unload_reps_all_vport           |    __unload_reps_all_vport     |
+--------------------------------------+--------------------------------+

These two flows will try to unload the same rep. Per original design,
once one flow unloads the rep, the state moves to REGISTERED. The 2nd
flow will no longer needs to do the unload and bails out. However, as
read and write of the state is not atomic, when 1st flow is doing the
unload, the state is still LOADED, 2nd flow is able to do the same
unload action. Kernel crash will happen.

To solve this, driver should do atomic test-and-set for the state. So
that only one flow can change the rep state from LOADED to REGISTERED,
and proceed to do the actual unloading.

Since the state is changing to atomic type, all other read/write should
be atomic action as well.

Fixes: f121e0ea9586 (net/mlx5: E-Switch, Add state to eswitch vport representors)
Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Vu Pham <vuhuong@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 .../mellanox/mlx5/core/eswitch_offloads.c     | 36 +++++++++----------
 include/linux/mlx5/eswitch.h                  |  2 +-
 2 files changed, 18 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 9b2d78ee22b88..d2d8da133082c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -363,7 +363,7 @@ static int esw_set_global_vlan_pop(struct mlx5_eswitch *esw, u8 val)
 	esw_debug(esw->dev, "%s applying global %s policy\n", __func__, val ? "pop" : "none");
 	for (vf_vport = 1; vf_vport < esw->enabled_vports; vf_vport++) {
 		rep = &esw->offloads.vport_reps[vf_vport];
-		if (rep->rep_if[REP_ETH].state != REP_LOADED)
+		if (atomic_read(&rep->rep_if[REP_ETH].state) != REP_LOADED)
 			continue;
 
 		err = __mlx5_eswitch_set_vport_vlan(esw, rep->vport, 0, 0, val);
@@ -1306,7 +1306,8 @@ int esw_offloads_init_reps(struct mlx5_eswitch *esw)
 		ether_addr_copy(rep->hw_id, hw_id);
 
 		for (rep_type = 0; rep_type < NUM_REP_TYPES; rep_type++)
-			rep->rep_if[rep_type].state = REP_UNREGISTERED;
+			atomic_set(&rep->rep_if[rep_type].state,
+				   REP_UNREGISTERED);
 	}
 
 	return 0;
@@ -1315,11 +1316,9 @@ int esw_offloads_init_reps(struct mlx5_eswitch *esw)
 static void __esw_offloads_unload_rep(struct mlx5_eswitch *esw,
 				      struct mlx5_eswitch_rep *rep, u8 rep_type)
 {
-	if (rep->rep_if[rep_type].state != REP_LOADED)
-		return;
-
-	rep->rep_if[rep_type].unload(rep);
-	rep->rep_if[rep_type].state = REP_REGISTERED;
+	if (atomic_cmpxchg(&rep->rep_if[rep_type].state,
+			   REP_LOADED, REP_REGISTERED) == REP_LOADED)
+		rep->rep_if[rep_type].unload(rep);
 }
 
 static void __unload_reps_special_vport(struct mlx5_eswitch *esw, u8 rep_type)
@@ -1380,16 +1379,15 @@ static int __esw_offloads_load_rep(struct mlx5_eswitch *esw,
 {
 	int err = 0;
 
-	if (rep->rep_if[rep_type].state != REP_REGISTERED)
-		return 0;
-
-	err = rep->rep_if[rep_type].load(esw->dev, rep);
-	if (err)
-		return err;
-
-	rep->rep_if[rep_type].state = REP_LOADED;
+	if (atomic_cmpxchg(&rep->rep_if[rep_type].state,
+			   REP_REGISTERED, REP_LOADED) == REP_REGISTERED) {
+		err = rep->rep_if[rep_type].load(esw->dev, rep);
+		if (err)
+			atomic_set(&rep->rep_if[rep_type].state,
+				   REP_REGISTERED);
+	}
 
-	return 0;
+	return err;
 }
 
 static int __load_reps_special_vport(struct mlx5_eswitch *esw, u8 rep_type)
@@ -2076,7 +2074,7 @@ void mlx5_eswitch_register_vport_reps(struct mlx5_eswitch *esw,
 		rep_if->get_proto_dev = __rep_if->get_proto_dev;
 		rep_if->priv = __rep_if->priv;
 
-		rep_if->state = REP_REGISTERED;
+		atomic_set(&rep_if->state, REP_REGISTERED);
 	}
 }
 EXPORT_SYMBOL(mlx5_eswitch_register_vport_reps);
@@ -2091,7 +2089,7 @@ void mlx5_eswitch_unregister_vport_reps(struct mlx5_eswitch *esw, u8 rep_type)
 		__unload_reps_all_vport(esw, max_vf, rep_type);
 
 	mlx5_esw_for_all_reps(esw, i, rep)
-		rep->rep_if[rep_type].state = REP_UNREGISTERED;
+		atomic_set(&rep->rep_if[rep_type].state, REP_UNREGISTERED);
 }
 EXPORT_SYMBOL(mlx5_eswitch_unregister_vport_reps);
 
@@ -2111,7 +2109,7 @@ void *mlx5_eswitch_get_proto_dev(struct mlx5_eswitch *esw,
 
 	rep = mlx5_eswitch_get_rep(esw, vport);
 
-	if (rep->rep_if[rep_type].state == REP_LOADED &&
+	if (atomic_read(&rep->rep_if[rep_type].state) == REP_LOADED &&
 	    rep->rep_if[rep_type].get_proto_dev)
 		return rep->rep_if[rep_type].get_proto_dev(rep);
 	return NULL;
diff --git a/include/linux/mlx5/eswitch.h b/include/linux/mlx5/eswitch.h
index 96d8435421de8..0ca77dd1429c0 100644
--- a/include/linux/mlx5/eswitch.h
+++ b/include/linux/mlx5/eswitch.h
@@ -35,7 +35,7 @@ struct mlx5_eswitch_rep_if {
 	void		       (*unload)(struct mlx5_eswitch_rep *rep);
 	void		       *(*get_proto_dev)(struct mlx5_eswitch_rep *rep);
 	void			*priv;
-	u8			state;
+	atomic_t		state;
 };
 
 struct mlx5_eswitch_rep {
-- 
2.20.1


  parent reply	other threads:[~2019-05-22 19:22 UTC|newest]

Thread overview: 88+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-05-22 19:15 [PATCH AUTOSEL 5.1 001/375] gfs2: Fix lru_count going negative Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 002/375] cxgb4: Fix error path in cxgb4_init_module Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 003/375] afs: Fix getting the afs.fid xattr Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 004/375] NFS: make nfs_match_client killable Sasha Levin
2019-05-23 15:02   ` Benjamin Coddington
2019-05-29 18:46     ` Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 005/375] gfs2: fix race between gfs2_freeze_func and unmount Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 006/375] io_uring: use cpu_online() to check p->sq_thread_cpu instead of cpu_possible() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 007/375] IB/hfi1: Fix WQ_MEM_RECLAIM warning Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 008/375] gfs2: Fix occasional glock use-after-free Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 009/375] mmc: core: Verify SD bus width Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 010/375] batman-adv: mcast: fix multicast tt/tvlv worker locking Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 011/375] ip6: fix skb leak in ip6frag_expire_frag_queue() Sasha Levin
2019-05-23  7:47   ` Stefan Bader
2019-05-29 18:48     ` Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 012/375] tools/bpf: fix perf build error with uClibc (seen on ARC) Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 013/375] i40e: Fix of memory leak and integer truncation in i40e_virtchnl.c Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 014/375] libbpf: fix invalid munmap call Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 015/375] selftests/bpf: set RLIMIT_MEMLOCK properly for test_libbpf_open.c Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 016/375] bpftool: exclude bash-completion/bpftool from .gitignore pattern Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 017/375] ice: Separate if conditions for ice_set_features() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 018/375] ice: Preserve VLAN Rx stripping settings Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 019/375] blk-mq: split blk_mq_alloc_and_init_hctx into two parts Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 020/375] blk-mq: grab .q_usage_counter when queuing request from plug code path Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 021/375] dmaengine: tegra210-dma: free dma controller in remove() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 022/375] net: ena: gcc 8: fix compilation warning Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 023/375] net: ena: fix: set freed objects to NULL to avoid failing future allocations Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 024/375] hv_netvsc: fix race that may miss tx queue wakeup Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 025/375] orangefs: truncate before updating size Sasha Levin
2019-05-22 21:44   ` martin
2019-05-29 18:51     ` Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 026/375] Bluetooth: Ignore CC events not matching the last HCI command Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 027/375] pinctrl: zte: fix leaked of_node references Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 028/375] ASoC: Intel: kbl_da7219_max98357a: Map BTN_0 to KEY_PLAYPAUSE Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 029/375] usb: dwc2: gadget: Increase descriptors count for ISOC's Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 030/375] usb: dwc3: move synchronize_irq() out of the spinlock protected block Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 031/375] usb: gadget: f_fs: don't free buffer prematurely Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 032/375] ASoC: hdmi-codec: unlock the device on startup errors Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 033/375] leds: avoid races with workqueue Sasha Levin
2019-05-24 22:55   ` Pavel Machek
2019-05-29 18:51     ` Sasha Levin
2019-06-17 15:57       ` Pavel Machek
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 034/375] powerpc/perf: Return accordingly on invalid chip-id in Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 035/375] powerpc/boot: Fix missing check of lseek() return value Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 036/375] powerpc/perf: Fix loop exit condition in nest_imc_event_init Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 037/375] spi: atmel-quadspi: fix crash while suspending Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 038/375] ASoC: imx: fix fiq dependencies Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 039/375] spi: pxa2xx: fix SCR (divisor) calculation Sasha Levin
2019-05-22 19:15 ` Sasha Levin [this message]
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 041/375] brcm80211: potential NULL dereference in brcmf_cfg80211_vndr_cmds_dcmd_handler() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 042/375] ACPI / property: fix handling of data_nodes in acpi_get_next_subnode() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 043/375] drm/nouveau/bar/nv50: ensure BAR is mapped Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 044/375] media: stm32-dcmi: return appropriate error codes during probe Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 045/375] ARM: vdso: Remove dependency with the arch_timer driver internals Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 046/375] arm64: Fix compiler warning from pte_unmap() with -Wunused-but-set-variable Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 047/375] mt76: remove mt76_queue dependency from tx_queue_skb function pointer Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 048/375] x86/ftrace: Set trampoline pages as executable Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 049/375] powerpc/watchdog: Use hrtimers for per-CPU heartbeat Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 050/375] cpufreq: Fix kobject memleak Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 051/375] scsi: qla2xxx: Fix a qla24xx_enable_msix() error path Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 052/375] scsi: qla2xxx: Fix abort handling in tcm_qla2xxx_write_pending() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 053/375] scsi: qla2xxx: Avoid that lockdep complains about unsafe locking in tcm_qla2xxx_close_session() Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 054/375] scsi: qla2xxx: Fix hardirq-unsafe locking Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 055/375] x86/modules: Avoid breaking W^X while loading modules Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 056/375] Btrfs: fix data bytes_may_use underflow with fallocate due to failed quota reserve Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 057/375] btrfs: fix panic during relocation after ENOSPC before writeback happens Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 058/375] btrfs: reloc: Fix NULL pointer dereference due to expanded reloc_root lifespan Sasha Levin
2019-05-22 19:15 ` [PATCH AUTOSEL 5.1 059/375] btrfs: Don't panic when we can't find a root key Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 060/375] iwlwifi: pcie: don't crash on invalid RX interrupt Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 061/375] rtc: 88pm860x: prevent use-after-free on device remove Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 062/375] rtc: stm32: manage the get_irq probe defer case Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 063/375] scsi: qedi: Abort ep termination if offload not scheduled Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 064/375] s390/kexec_file: Fix detection of text segment in ELF loader Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 065/375] ALSA: hda: fix unregister device twice on ASoC driver Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 066/375] sched/nohz: Run NOHZ idle load balancer on HK_FLAG_MISC CPUs Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 067/375] net: ethernet: ti: cpsw: fix allmulti cfg in dual_mac mode Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 068/375] w1: fix the resume command API Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 069/375] net: hns3: fix pause configure fail problem Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 070/375] net: hns3: fix for TX clean num when cleaning TX BD Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 071/375] net: phy: improve genphy_soft_reset Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 072/375] s390: qeth: address type mismatch warning Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 073/375] arm64: futex: Fix FUTEX_WAKE_OP atomic ops with non-zero result value Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 074/375] net: hns3: use atomic_t replace u32 for arq's count Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 075/375] dmaengine: pl330: _stop: clear interrupt status Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 076/375] mac80211/cfg80211: update bss channel on channel switch Sasha Levin
2019-05-22 19:16 ` [PATCH AUTOSEL 5.1 077/375] USB: serial: fix initial-termios handling Sasha Levin
2019-05-23  5:26   ` Johan Hovold
2019-05-29 18:56     ` Sasha Levin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20190522192115.22666-40-sashal@kernel.org \
    --to=sashal@kernel.org \
    --cc=bodong@mellanox.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=parav@mellanox.com \
    --cc=saeedm@mellanox.com \
    --cc=stable@vger.kernel.org \
    --cc=vuhuong@mellanox.com \
    --subject='Re: [PATCH AUTOSEL 5.1 040/375] net/mlx5: E-Switch, Use atomic rep state to serialize state change' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).