public inbox for stable@vger.kernel.org
 help / color / mirror / Atom feed
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Scott Wood <swood@redhat.com>,
	Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
	"Paul E . McKenney" <paulmck@kernel.org>,
	Sasha Levin <sashal@kernel.org>,
	dave@stgolabs.net, josh@joshtriplett.org, rcu@vger.kernel.org
Subject: [PATCH AUTOSEL 5.15 009/146] rcutorture: Avoid problematic critical section nesting on PREEMPT_RT
Date: Mon,  8 Nov 2021 12:42:36 -0500	[thread overview]
Message-ID: <20211108174453.1187052-9-sashal@kernel.org> (raw)
In-Reply-To: <20211108174453.1187052-1-sashal@kernel.org>

From: Scott Wood <swood@redhat.com>

[ Upstream commit 71921a9606ddbcc1d98c00eca7ae82c373d1fecd ]

rcutorture is generating some nesting scenarios that are not compatible on PREEMPT_RT.
For example:
	preempt_disable();
	rcu_read_lock_bh();
	preempt_enable();
	rcu_read_unlock_bh();

The problem here is that on PREEMPT_RT the bottom halves have to be
disabled and enabled in preemptible context.

Reorder locking: start with BH locking and continue with then with
disabling preemption or interrupts. In the unlocking do it reverse by
first enabling interrupts and preemption and BH at the very end.
Ensure that on PREEMPT_RT BH locking remains unchanged if in
non-preemptible context.

Link: https://lkml.kernel.org/r/20190911165729.11178-6-swood@redhat.com
Link: https://lkml.kernel.org/r/20210819182035.GF4126399@paulmck-ThinkPad-P17-Gen-1
Signed-off-by: Scott Wood <swood@redhat.com>
[bigeasy: Drop ATOM_BH, make it only about changing BH in atomic
context. Allow enabling RCU in IRQ-off section. Reword commit message.]
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 kernel/rcu/rcutorture.c | 48 ++++++++++++++++++++++++++++++-----------
 1 file changed, 36 insertions(+), 12 deletions(-)

diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index ab4215266ebee..968696ace8f3f 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -1432,28 +1432,34 @@ static void rcutorture_one_extend(int *readstate, int newstate,
 	/* First, put new protection in place to avoid critical-section gap. */
 	if (statesnew & RCUTORTURE_RDR_BH)
 		local_bh_disable();
+	if (statesnew & RCUTORTURE_RDR_RBH)
+		rcu_read_lock_bh();
 	if (statesnew & RCUTORTURE_RDR_IRQ)
 		local_irq_disable();
 	if (statesnew & RCUTORTURE_RDR_PREEMPT)
 		preempt_disable();
-	if (statesnew & RCUTORTURE_RDR_RBH)
-		rcu_read_lock_bh();
 	if (statesnew & RCUTORTURE_RDR_SCHED)
 		rcu_read_lock_sched();
 	if (statesnew & RCUTORTURE_RDR_RCU)
 		idxnew = cur_ops->readlock() << RCUTORTURE_RDR_SHIFT;
 
-	/* Next, remove old protection, irq first due to bh conflict. */
+	/*
+	 * Next, remove old protection, in decreasing order of strength
+	 * to avoid unlock paths that aren't safe in the stronger
+	 * context. Namely: BH can not be enabled with disabled interrupts.
+	 * Additionally PREEMPT_RT requires that BH is enabled in preemptible
+	 * context.
+	 */
 	if (statesold & RCUTORTURE_RDR_IRQ)
 		local_irq_enable();
-	if (statesold & RCUTORTURE_RDR_BH)
-		local_bh_enable();
 	if (statesold & RCUTORTURE_RDR_PREEMPT)
 		preempt_enable();
-	if (statesold & RCUTORTURE_RDR_RBH)
-		rcu_read_unlock_bh();
 	if (statesold & RCUTORTURE_RDR_SCHED)
 		rcu_read_unlock_sched();
+	if (statesold & RCUTORTURE_RDR_BH)
+		local_bh_enable();
+	if (statesold & RCUTORTURE_RDR_RBH)
+		rcu_read_unlock_bh();
 	if (statesold & RCUTORTURE_RDR_RCU) {
 		bool lockit = !statesnew && !(torture_random(trsp) & 0xffff);
 
@@ -1496,6 +1502,9 @@ rcutorture_extend_mask(int oldmask, struct torture_random_state *trsp)
 	int mask = rcutorture_extend_mask_max();
 	unsigned long randmask1 = torture_random(trsp) >> 8;
 	unsigned long randmask2 = randmask1 >> 3;
+	unsigned long preempts = RCUTORTURE_RDR_PREEMPT | RCUTORTURE_RDR_SCHED;
+	unsigned long preempts_irq = preempts | RCUTORTURE_RDR_IRQ;
+	unsigned long bhs = RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH;
 
 	WARN_ON_ONCE(mask >> RCUTORTURE_RDR_SHIFT);
 	/* Mostly only one bit (need preemption!), sometimes lots of bits. */
@@ -1503,11 +1512,26 @@ rcutorture_extend_mask(int oldmask, struct torture_random_state *trsp)
 		mask = mask & randmask2;
 	else
 		mask = mask & (1 << (randmask2 % RCUTORTURE_RDR_NBITS));
-	/* Can't enable bh w/irq disabled. */
-	if ((mask & RCUTORTURE_RDR_IRQ) &&
-	    ((!(mask & RCUTORTURE_RDR_BH) && (oldmask & RCUTORTURE_RDR_BH)) ||
-	     (!(mask & RCUTORTURE_RDR_RBH) && (oldmask & RCUTORTURE_RDR_RBH))))
-		mask |= RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH;
+
+	/*
+	 * Can't enable bh w/irq disabled.
+	 */
+	if (mask & RCUTORTURE_RDR_IRQ)
+		mask |= oldmask & bhs;
+
+	/*
+	 * Ideally these sequences would be detected in debug builds
+	 * (regardless of RT), but until then don't stop testing
+	 * them on non-RT.
+	 */
+	if (IS_ENABLED(CONFIG_PREEMPT_RT)) {
+		/* Can't modify BH in atomic context */
+		if (oldmask & preempts_irq)
+			mask &= ~bhs;
+		if ((oldmask | mask) & preempts_irq)
+			mask |= oldmask & bhs;
+	}
+
 	return mask ?: RCUTORTURE_RDR_RCU;
 }
 
-- 
2.33.0


  parent reply	other threads:[~2021-11-09  0:59 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-08 17:42 [PATCH AUTOSEL 5.15 001/146] dma-buf: WARN on dmabuf release with pending attachments Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 002/146] drm: panel-orientation-quirks: Update the Lenovo Ideapad D330 quirk (v2) Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 003/146] drm: panel-orientation-quirks: Add quirk for KD Kurio Smart C15200 2-in-1 Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 004/146] drm: panel-orientation-quirks: Add quirk for the Samsung Galaxy Book 10.6 Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 005/146] Bluetooth: sco: Fix lock_sock() blockage by memcpy_from_msg() Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 006/146] Bluetooth: fix use-after-free error in lock_sock_nested() Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 007/146] Bluetooth: call sock_hold earlier in sco_conn_del Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 008/146] drm/panel-orientation-quirks: add Valve Steam Deck Sasha Levin
2021-11-08 17:42 ` Sasha Levin [this message]
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 010/146] platform/x86: wmi: do not fail if disabling fails Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 011/146] drm/amdgpu: move iommu_resume before ip init/resume Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 012/146] MIPS: lantiq: dma: add small delay after reset Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 013/146] MIPS: lantiq: dma: reset correct number of channel Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 014/146] locking/lockdep: Avoid RCU-induced noinstr fail Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 015/146] net: sched: update default qdisc visibility after Tx queue cnt changes Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 016/146] ACPI: resources: Add DMI-based legacy IRQ override quirk Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 017/146] rcu-tasks: Move RTGS_WAIT_CBS to beginning of rcu_tasks_kthread() loop Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 018/146] smackfs: Fix use-after-free in netlbl_catmap_walk() Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 019/146] ath11k: Align bss_chan_info structure with firmware Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 020/146] crypto: aesni - check walk.nbytes instead of err Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 021/146] NET: IPV4: fix error "do not initialise globals to 0" Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 022/146] x86/mm/64: Improve stack overflow warnings Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 023/146] x86: Increase exception stack sizes Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 024/146] mwifiex: Run SET_BSS_MODE when changing from P2P to STATION vif-type Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 025/146] mwifiex: Properly initialize private structure on interface type changes Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 026/146] spi: Check we have a spi_device_id for each DT compatible Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 027/146] fscrypt: allow 256-bit master keys with AES-256-XTS Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 028/146] drm/amdgpu: Fix MMIO access page fault Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 029/146] drm/amd/display: Fix null pointer dereference for encoders Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 030/146] crypto: api - Fix built-in testing dependency failures Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 031/146] selftests: net: fib_nexthops: Wait before checking reported idle time Sasha Levin
2021-11-08 17:42 ` [PATCH AUTOSEL 5.15 032/146] leds: trigger: use RCU to protect the led_cdevs list Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 033/146] ath11k: Avoid reg rules update during firmware recovery Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 034/146] ath11k: add handler for scan event WMI_SCAN_EVENT_DEQUEUED Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 035/146] ath11k: Change DMA_FROM_DEVICE to DMA_TO_DEVICE when map reinjected packets Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 036/146] ath10k: high latency fixes for beacon buffer Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 037/146] octeontx2-pf: Enable promisc/allmulti match MCAM entries Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 038/146] media: mt9p031: Fix corrupted frame after restarting stream Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 039/146] media: netup_unidvb: handle interrupt properly according to the firmware Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 040/146] media: atomisp: Fix error handling in probe Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 041/146] media: stm32: Potential NULL pointer dereference in dcmi_irq_thread() Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 042/146] media: uvcvideo: Set capability in s_param Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 043/146] media: uvcvideo: Return -EIO for control errors Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 044/146] media: uvcvideo: Set unique vdev name based in type Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 045/146] media: vidtv: Fix memory leak in remove Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 046/146] media: s5p-mfc: fix possible null-pointer dereference in s5p_mfc_probe() Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 047/146] media: s5p-mfc: Add checking to s5p_mfc_probe() Sasha Levin
2021-11-08 17:43 ` [PATCH AUTOSEL 5.15 048/146] media: videobuf2: rework vb2_mem_ops API Sasha Levin
2021-11-09  7:54 ` AUTOSEL series truncated was -- Re: [PATCH AUTOSEL 5.15 001/146] dma-buf: WARN on dmabuf release with pending attachments Pavel Machek
2021-11-09  8:07   ` Christian König
2021-11-09 16:05   ` Randy Dunlap
2021-11-14 14:25     ` 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=20211108174453.1187052-9-sashal@kernel.org \
    --to=sashal@kernel.org \
    --cc=bigeasy@linutronix.de \
    --cc=dave@stgolabs.net \
    --cc=josh@joshtriplett.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=paulmck@kernel.org \
    --cc=rcu@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=swood@redhat.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox