From: Tony Lindgren <tony@atomide.com>
To: i2c@lm-sensors.org
Cc: linux-omap@vger.kernel.org, Paul Walmsley <paul@pwsan.com>,
Tony Lindgren <tony@atomide.com>
Subject: [PATCH 2/8] i2c-omap: Close suspected race between omap_i2c_idle() and omap_i2c_isr()
Date: Thu, 25 Sep 2008 10:53:48 +0300 [thread overview]
Message-ID: <1222329234-31473-3-git-send-email-tony@atomide.com> (raw)
In-Reply-To: <1222329234-31473-2-git-send-email-tony@atomide.com>
From: Paul Walmsley <paul@pwsan.com>
omap_i2c_idle() sets an internal flag, "dev->idle", instructing its
ISR to decline interrupts. It sets this flag before it actually masks
the interrupts on the I2C controller. This is problematic, since an
I2C interrupt could arrive after dev->idle is set, but before the
interrupt source is masked. When this happens, Linux disables the I2C
controller's IRQ, causing all future transactions on the bus to fail.
Symptoms, happening on about 7% of boots:
irq 56: nobody cared (try booting with the "irqpoll" option)
<warning traceback here>
Disabling IRQ #56
i2c_omap i2c_omap.1: controller timed out
In omap_i2c_idle(), this patch sets dev->idle only after the interrupt
mask write to the I2C controller has left the ARM write buffer.
That's probably the major offender. For additional prophylaxis, in
omap_i2c_unidle(), the patch clears the dev->idle flag before
interrupts are enabled, rather than afterwards.
The patch has survived twenty-two reboots on the 3430SDP here without
wedging I2C1. Not absolutely dispositive, but promising!
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
drivers/i2c/busses/i2c-omap.c | 10 ++++++++--
1 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index cfb76f5..a06ad42 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -181,22 +181,28 @@ static void omap_i2c_unidle(struct omap_i2c_dev *dev)
if (dev->iclk != NULL)
clk_enable(dev->iclk);
clk_enable(dev->fclk);
+ dev->idle = 0;
if (dev->iestate)
omap_i2c_write_reg(dev, OMAP_I2C_IE_REG, dev->iestate);
- dev->idle = 0;
}
static void omap_i2c_idle(struct omap_i2c_dev *dev)
{
u16 iv;
- dev->idle = 1;
dev->iestate = omap_i2c_read_reg(dev, OMAP_I2C_IE_REG);
omap_i2c_write_reg(dev, OMAP_I2C_IE_REG, 0);
if (dev->rev1)
iv = omap_i2c_read_reg(dev, OMAP_I2C_IV_REG); /* Read clears */
else
omap_i2c_write_reg(dev, OMAP_I2C_STAT_REG, dev->iestate);
+ /*
+ * The wmb() is to ensure that the I2C interrupt mask write
+ * reaches the I2C controller before the dev->idle store
+ * occurs.
+ */
+ wmb();
+ dev->idle = 1;
clk_disable(dev->fclk);
if (dev->iclk != NULL)
clk_disable(dev->iclk);
--
1.5.6.rc3.21.g8c6b5
next prev parent reply other threads:[~2008-09-25 7:54 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-09-25 7:53 [PATCH 0/8] Updates for i2c-omap from linux-omap tree for review, v2 Tony Lindgren
2008-09-25 7:53 ` [PATCH 1/8] i2c-omap: Do not use interruptible wait call in omap_i2c_xfer_msg Tony Lindgren
2008-09-25 7:53 ` Tony Lindgren [this message]
2008-09-25 7:53 ` [PATCH 3/8] i2c-omap: Add high-speed support to omap-i2c Tony Lindgren
2008-09-25 7:53 ` [PATCH 4/8] i2c-omap: FIFO handling support and broken hw workaround for i2c-omap Tony Lindgren
2008-09-25 7:53 ` [PATCH 5/8] i2c-omap: Add support on 34xx Tony Lindgren
2008-09-25 7:53 ` [PATCH 6/8] i2c-omap: Mark init-only functions as __init Tony Lindgren
2008-09-25 7:53 ` [PATCH 7/8] i2c-omap: Don't compile in OMAP15xx I2C ISR for non-OMAP15xx builds Tony Lindgren
2008-09-25 7:53 ` [PATCH 8/8] i2c-omap: Clean-up i2c-omap Tony Lindgren
2008-09-25 11:40 ` [PATCH 8/8] i2c-omap: Clean-up i2c-omap, v3 Tony Lindgren
2008-09-29 22:31 ` [i2c] [PATCH 8/8] i2c-omap: Clean-up i2c-omap Ben Dooks
2008-10-17 15:41 ` Tony Lindgren
2008-09-29 22:29 ` [i2c] [PATCH 7/8] i2c-omap: Don't compile in OMAP15xx I2C ISR for non-OMAP15xx builds Ben Dooks
2008-09-29 23:06 ` David Brownell
2008-09-30 5:35 ` Paul Walmsley
2008-09-30 5:36 ` Paul Walmsley
2008-10-17 15:40 ` Tony Lindgren
2008-09-29 22:30 ` [i2c] [PATCH 6/8] i2c-omap: Mark init-only functions as __init Ben Dooks
2008-10-17 15:39 ` Tony Lindgren
2008-10-17 17:07 ` Tony Lindgren
2008-09-29 22:28 ` [i2c] [PATCH 5/8] i2c-omap: Add support on 34xx Ben Dooks
2008-10-17 15:38 ` Tony Lindgren
2008-09-25 11:39 ` [PATCH 4/8] i2c-omap: FIFO handling support and broken hw workaround for i2c-omap, v3 Tony Lindgren
2008-09-29 22:27 ` [i2c] [PATCH 4/8] i2c-omap: FIFO handling support and broken hw workaround for i2c-omap Ben Dooks
2008-09-29 22:25 ` [i2c] [PATCH 3/8] i2c-omap: Add high-speed support to omap-i2c Ben Dooks
2008-10-17 15:37 ` Tony Lindgren
2008-09-29 22:21 ` [i2c] [PATCH 2/8] i2c-omap: Close suspected race between omap_i2c_idle() and omap_i2c_isr() Ben Dooks
2008-09-29 22:21 ` [i2c] [PATCH 1/8] i2c-omap: Do not use interruptible wait call in omap_i2c_xfer_msg Ben Dooks
2008-09-30 11:51 ` Jarkko Nikula
2008-10-17 15:35 ` Tony Lindgren
2008-09-29 22:23 ` [i2c] [PATCH 0/8] Updates for i2c-omap from linux-omap tree for review, v2 Ben Dooks
2008-09-30 5:35 ` Tony Lindgren
2008-10-17 16:04 ` git-pull request for i2c-omap changes (Re: [i2c] [PATCH 0/8] Updates for i2c-omap from linux-omap tree for review, v2) Tony Lindgren
-- strict thread matches above, loose matches on Subject: below --
2008-09-25 7:16 [PATCH 0/8] Updates for i2c-omap from linux-omap tree for review Tony Lindgren
2008-09-25 7:16 ` [PATCH 1/8] i2c-omap: Do not use interruptible wait call in omap_i2c_xfer_msg Tony Lindgren
2008-09-25 7:16 ` [PATCH 2/8] i2c-omap: Close suspected race between omap_i2c_idle() and omap_i2c_isr() Tony Lindgren
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=1222329234-31473-3-git-send-email-tony@atomide.com \
--to=tony@atomide.com \
--cc=i2c@lm-sensors.org \
--cc=linux-omap@vger.kernel.org \
--cc=paul@pwsan.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).