linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
From: Rohan McLure <rmclure@linux.ibm.com>
To: linuxppc-dev@lists.ozlabs.org
Cc: Rohan McLure <rmclure@ibm.com>,
	arnd@arndb.de, npiggin@gmail.com,
	Rohan McLure <rmclure@linux.ibm.com>
Subject: [PATCH 07/12] powerpc: powernv: Fix KCSAN datarace warnings on idle_state contention
Date: Mon,  8 May 2023 12:01:15 +1000	[thread overview]
Message-ID: <20230508020120.218494-8-rmclure@linux.ibm.com> (raw)
In-Reply-To: <20230508020120.218494-1-rmclure@linux.ibm.com>

The idle_state entry in the PACA on PowerNV features a bit which is
atomically tested and set through ldarx/stdcx. to be used as a spinlock.
This lock then guards access to other bit fields of idle_state. KCSAN
cannot differentiate between any of these bitfield accesses as they all
are implemented by 8-byte store/load instructions, thus cores contending
on the bit-lock appear to data race with modifications to idle_state.

Separate the bit-lock entry from the data guarded by the lock to avoid
the possibility of data races being detected by KCSAN.

Suggested-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Rohan McLure <rmclure@ibm.com>
---
 arch/powerpc/include/asm/paca.h       |  1 +
 arch/powerpc/platforms/powernv/idle.c | 20 +++++++++++---------
 2 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/arch/powerpc/include/asm/paca.h b/arch/powerpc/include/asm/paca.h
index da0377f46597..cb325938766a 100644
--- a/arch/powerpc/include/asm/paca.h
+++ b/arch/powerpc/include/asm/paca.h
@@ -191,6 +191,7 @@ struct paca_struct {
 #ifdef CONFIG_PPC_POWERNV
 	/* PowerNV idle fields */
 	/* PNV_CORE_IDLE_* bits, all siblings work on thread 0 paca */
+	unsigned long idle_lock; /* A value of 1 means acquired */
 	unsigned long idle_state;
 	union {
 		/* P7/P8 specific fields */
diff --git a/arch/powerpc/platforms/powernv/idle.c b/arch/powerpc/platforms/powernv/idle.c
index 841cb7f31f4f..97dbb7bc2b00 100644
--- a/arch/powerpc/platforms/powernv/idle.c
+++ b/arch/powerpc/platforms/powernv/idle.c
@@ -246,9 +246,9 @@ static inline void atomic_lock_thread_idle(void)
 {
 	int cpu = raw_smp_processor_id();
 	int first = cpu_first_thread_sibling(cpu);
-	unsigned long *state = &paca_ptrs[first]->idle_state;
+	unsigned long *lock = &paca_ptrs[first]->idle_lock;
 
-	while (unlikely(test_and_set_bit_lock(NR_PNV_CORE_IDLE_LOCK_BIT, state)))
+	while (unlikely(test_and_set_bit_lock(NR_PNV_CORE_IDLE_LOCK_BIT, lock)))
 		barrier();
 }
 
@@ -258,29 +258,31 @@ static inline void atomic_unlock_and_stop_thread_idle(void)
 	int first = cpu_first_thread_sibling(cpu);
 	unsigned long thread = 1UL << cpu_thread_in_core(cpu);
 	unsigned long *state = &paca_ptrs[first]->idle_state;
+	unsigned long *lock = &paca_ptrs[first]->idle_lock;
 	u64 s = READ_ONCE(*state);
 	u64 new, tmp;
 
-	BUG_ON(!(s & PNV_CORE_IDLE_LOCK_BIT));
+	BUG_ON(!(READ_ONCE(*lock) & PNV_CORE_IDLE_LOCK_BIT));
 	BUG_ON(s & thread);
 
 again:
-	new = (s | thread) & ~PNV_CORE_IDLE_LOCK_BIT;
+	new = s | thread;
 	tmp = cmpxchg(state, s, new);
 	if (unlikely(tmp != s)) {
 		s = tmp;
 		goto again;
 	}
+	clear_bit_unlock(NR_PNV_CORE_IDLE_LOCK_BIT, lock);
 }
 
 static inline void atomic_unlock_thread_idle(void)
 {
 	int cpu = raw_smp_processor_id();
 	int first = cpu_first_thread_sibling(cpu);
-	unsigned long *state = &paca_ptrs[first]->idle_state;
+	unsigned long *lock = &paca_ptrs[first]->idle_lock;
 
-	BUG_ON(!test_bit(NR_PNV_CORE_IDLE_LOCK_BIT, state));
-	clear_bit_unlock(NR_PNV_CORE_IDLE_LOCK_BIT, state);
+	BUG_ON(!test_bit(NR_PNV_CORE_IDLE_LOCK_BIT, lock));
+	clear_bit_unlock(NR_PNV_CORE_IDLE_LOCK_BIT, lock);
 }
 
 /* P7 and P8 */
@@ -380,9 +382,9 @@ static unsigned long power7_idle_insn(unsigned long type)
 		sprs.uamor	= mfspr(SPRN_UAMOR);
 	}
 
-	local_paca->thread_idle_state = type;
+	WRITE_ONCE(local_paca->thread_idle_state, type);
 	srr1 = isa206_idle_insn_mayloss(type);		/* go idle */
-	local_paca->thread_idle_state = PNV_THREAD_RUNNING;
+	WRITE_ONCE(local_paca->thread_idle_state, PNV_THREAD_RUNNING);
 
 	WARN_ON_ONCE(!srr1);
 	WARN_ON_ONCE(mfmsr() & (MSR_IR|MSR_DR));
-- 
2.37.2


  parent reply	other threads:[~2023-05-08  2:05 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-08  2:01 [PATCH 00/12] powerpc: KCSAN fix warnings and mark accesses Rohan McLure
2023-05-08  2:01 ` [PATCH 01/12] powerpc: qspinlock: Fix qnode->locked value interpretation Rohan McLure
2023-05-09  2:01   ` Nicholas Piggin
2023-05-09  4:26     ` Rohan McLure
2023-05-08  2:01 ` [PATCH 02/12] powerpc: qspinlock: Mark accesses to qnode lock checks Rohan McLure
2023-05-09  2:02   ` Nicholas Piggin
2023-05-08  2:01 ` [PATCH 03/12] powerpc: qspinlock: Enforce qnode writes prior to publishing to queue Rohan McLure
2023-05-09  2:04   ` Nicholas Piggin
2023-05-09  5:26     ` Rohan McLure
2023-05-09  6:45       ` Nicholas Piggin
2023-05-08  2:01 ` [PATCH 04/12] asm-generic/mmiowb: Mark accesses to fix KCSAN warnings Rohan McLure
2023-05-08  6:30   ` Arnd Bergmann
2023-05-08 15:44   ` [PATCH 4/12] " Gautam Menghani
2023-05-09  2:16   ` [PATCH 04/12] " Nicholas Piggin
2023-05-08  2:01 ` [PATCH 05/12] powerpc: Mark [h]ssr_valid accesses in check_return_regs_valid Rohan McLure
2023-05-09  2:17   ` Nicholas Piggin
2023-05-08  2:01 ` [PATCH 06/12] powerpc: Mark accesses to power_save callback in arch_cpu_idle Rohan McLure
2023-05-09  2:21   ` Nicholas Piggin
2023-05-08  2:01 ` Rohan McLure [this message]
2023-05-09  2:26   ` [PATCH 07/12] powerpc: powernv: Fix KCSAN datarace warnings on idle_state contention Nicholas Piggin
2023-05-10  2:00     ` Rohan McLure
2023-05-08  2:01 ` [PATCH 08/12] powerpc: Annotate accesses to ipi message flags Rohan McLure
2023-05-09  2:28   ` Nicholas Piggin
2023-05-08  2:01 ` [PATCH 09/12] powerpc: Mark writes registering ipi to host cpu through kvm Rohan McLure
2023-05-09  2:30   ` Nicholas Piggin
2023-05-08  2:01 ` [PATCH 10/12] powerpc: powernv: Annotate data races in opal events Rohan McLure
2023-05-09  2:31   ` Nicholas Piggin
2023-05-08  2:01 ` [PATCH 11/12] powerpc: powernv: Annotate asynchronous access to opal tokens Rohan McLure
2023-05-08  2:01 ` [PATCH 12/12] powerpc: Mark asynchronous accesses to irq_data Rohan McLure

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=20230508020120.218494-8-rmclure@linux.ibm.com \
    --to=rmclure@linux.ibm.com \
    --cc=arnd@arndb.de \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=npiggin@gmail.com \
    --cc=rmclure@ibm.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;
as well as URLs for NNTP newsgroup(s).