linux-kernel - [PATCH 2/4] ide: Handle irq disabling consistently

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri,  4 May 2018 16:24:44 +0200
From:   Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To:     linux-ide@...r.kernel.org
Cc:     linux-kernel@...r.kernel.org, tglx@...utronix.de,
        "David S. Miller" <davem@...emloft.net>,
        Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Subject: [PATCH 2/4] ide: Handle irq disabling consistently

ide_timer_expiry() disables interrupt at function entry when acquiring
hwif->lock. Before disabling the device interrupt it unlocks hwif->lock,
but interrupts stay disabled. After the call to disable_irq() interrupts
are disabled again, which is a pointless exercise.

After the device irq handler has been invoked with interrupts disabled,
hwif->lock is acquired again with spin_lock_irq() because the device irq
handler might have reenabled interrupts. This is not documented and
confusing for the casual reader.

Remove the redundant local_irq_disable() and add a comment which explains
why hwif->lock has to be reacquired with spin_lock_irq().

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
---
 drivers/ide/ide-io.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c
index 6f25da56a169..a444bad7a2aa 100644
--- a/drivers/ide/ide-io.c
+++ b/drivers/ide/ide-io.c
@@ -659,8 +659,7 @@ void ide_timer_expiry (struct timer_list *t)
 		spin_unlock(&hwif->lock);
 		/* disable_irq_nosync ?? */
 		disable_irq(hwif->irq);
-		/* local CPU only, as if we were handling an interrupt */
-		local_irq_disable();
+
 		if (hwif->polling) {
 			startstop = handler(drive);
 		} else if (drive_is_ready(drive)) {
@@ -679,6 +678,7 @@ void ide_timer_expiry (struct timer_list *t)
 				startstop = ide_error(drive, "irq timeout",
 					hwif->tp_ops->read_status(hwif));
 		}
+		/* Disable interrupts again, `handler' might have enabled it */
 		spin_lock_irq(&hwif->lock);
 		enable_irq(hwif->irq);
 		if (startstop == ide_stopped && hwif->polling == 0) {
-- 
2.17.0