From: Geoff Levand <geoffrey.levand@am.sony.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: dwalker@mvista.com, matthias.kaehlcke@gmail.com,
linux-kernel@vger.kernel.org, linux@bohmer.net,
linuxppc-dev@ozlabs.org, mingo@elte.hu, kjwinchester@gmail.com
Subject: Re: [PATCH 1/3] ps3: vuart: fix error path locking
Date: Tue, 18 Dec 2007 19:04:58 -0800 [thread overview]
Message-ID: <47688A5A.1090902@am.sony.com> (raw)
In-Reply-To: <20071218171050.5bf220fa.akpm@linux-foundation.org>
On 12/18/2007 05:10 PM, Andrew Morton wrote:
> On Wed, 12 Dec 2007 18:00:12 -0800
> Geoff Levand <geoffrey.levand@am.sony.com> wrote:
>
>> > This stray down would cause a permanent sleep which doesn't seem correct.
>> > The other uses of this semaphore appear fairly mutex like it's even initialized
>> > with init_MUTEX() .. So here a patch for removing this one down().
>> >
>> > Signed-off-by: Daniel Walker <dwalker@mvista.com>
>> >
>> > ---
>> > drivers/ps3/ps3-vuart.c | 1 -
>> > 1 file changed, 1 deletion(-)
>>
>>
>> Signed-off-by: Geoff Levand <geoffrey.levand@am.sony.com>
>>
>>
>> Looks, good.
>
> Looks bad to me.
Hi Andrew,
Unfortunately there wasn't enough context in the patch to see
that there is a down() earlier in the routine, and that the patch
does indeed remove an incorrectly placed down(). Here is the
entire routine, marked with what the patch removes.
static int ps3_vuart_probe(struct ps3_system_bus_device *dev)
{
int result;
struct ps3_vuart_port_driver *drv;
struct ps3_vuart_port_priv *priv = NULL;
dev_dbg(&dev->core, "%s:%d\n", __func__, __LINE__);
drv = ps3_system_bus_dev_to_vuart_drv(dev);
dev_dbg(&dev->core, "%s:%d: (%s)\n", __func__, __LINE__,
drv->core.core.name);
BUG_ON(!drv);
if (dev->port_number >= PORT_COUNT) {
BUG();
return -EINVAL;
}
down(&vuart_bus_priv.probe_mutex);
result = ps3_vuart_bus_interrupt_get();
if (result)
goto fail_setup_interrupt;
if (vuart_bus_priv.devices[dev->port_number]) {
dev_dbg(&dev->core, "%s:%d: port busy (%d)\n", __func__,
__LINE__, dev->port_number);
result = -EBUSY;
goto fail_busy;
}
vuart_bus_priv.devices[dev->port_number] = dev;
/* Setup dev->driver_priv. */
dev->driver_priv = kzalloc(sizeof(struct ps3_vuart_port_priv),
GFP_KERNEL);
if (!dev->driver_priv) {
result = -ENOMEM;
goto fail_dev_malloc;
}
priv = to_port_priv(dev);
INIT_LIST_HEAD(&priv->tx_list.head);
spin_lock_init(&priv->tx_list.lock);
INIT_LIST_HEAD(&priv->rx_list.head);
spin_lock_init(&priv->rx_list.lock);
INIT_WORK(&priv->rx_list.work.work, NULL);
priv->rx_list.work.trigger = 0;
priv->rx_list.work.dev = dev;
/* clear stale pending interrupts */
ps3_vuart_clear_rx_bytes(dev, 0);
ps3_vuart_set_interrupt_mask(dev, INTERRUPT_MASK_RX);
ps3_vuart_set_triggers(dev, 1, 1);
if (drv->probe)
result = drv->probe(dev);
else {
result = 0;
dev_info(&dev->core, "%s:%d: no probe method\n", __func__,
__LINE__);
}
if (result) {
dev_dbg(&dev->core, "%s:%d: drv->probe failed\n",
__func__, __LINE__);
removed >>>>>> down(&vuart_bus_priv.probe_mutex); <<<<<<<<<<<
goto fail_probe;
}
up(&vuart_bus_priv.probe_mutex);
return result;
fail_probe:
ps3_vuart_set_interrupt_mask(dev, 0);
kfree(dev->driver_priv);
dev->driver_priv = NULL;
fail_dev_malloc:
vuart_bus_priv.devices[dev->port_number] = NULL;
fail_busy:
ps3_vuart_bus_interrupt_put();
fail_setup_interrupt:
up(&vuart_bus_priv.probe_mutex);
dev_dbg(&dev->core, "%s:%d: failed\n", __func__, __LINE__);
return result;
}
Thanks for taking the time to scrutinize.
-Geoff
next prev parent reply other threads:[~2007-12-19 3:30 UTC|newest]
Thread overview: 92+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20071213003023.117964080@mvista.com>
2007-12-12 8:00 ` [PATCH 1/3] ps3: vuart: fix error path locking Daniel Walker
2007-12-13 2:00 ` Geoff Levand
2007-12-19 1:10 ` Andrew Morton
2007-12-19 1:54 ` Daniel Walker
2007-12-19 3:04 ` Geoff Levand [this message]
2007-12-20 19:32 ` Daniel Walker
2007-12-20 20:06 ` Andrew Morton
2007-12-20 20:13 ` Daniel Walker
2007-12-12 8:00 ` [PATCH 2/3] driver base: memory: semaphore to mutex Daniel Walker
2007-12-12 8:00 ` [PATCH 3/3] net: wireless: bcm43xx: big_buffer_sem " Daniel Walker
2007-12-13 0:48 ` Michael Buesch
2007-12-13 1:17 ` Ray Lee
2007-12-13 13:45 ` Michael Buesch
2007-12-14 0:05 ` Ray Lee
2007-12-14 0:43 ` Michael Buesch
2007-12-14 0:55 ` Harvey Harrison
2007-12-14 10:45 ` Michael Buesch
2007-12-14 1:12 ` Ray Lee
2007-12-14 10:49 ` Michael Buesch
2007-12-14 11:15 ` Ingo Molnar
2007-12-14 11:39 ` Michael Buesch
2007-12-14 12:16 ` Ingo Molnar
2007-12-14 12:31 ` Michael Buesch
2007-12-14 12:53 ` Ingo Molnar
2007-12-14 13:08 ` Michael Buesch
2007-12-14 16:06 ` Ray Lee
2007-12-14 16:49 ` Michael Buesch
2007-12-14 18:01 ` Ray Lee
2007-12-14 18:11 ` Ingo Molnar
2007-12-14 18:47 ` Ray Lee
2007-12-14 18:22 ` Michael Buesch
2007-12-14 18:45 ` Ray Lee
2007-12-14 19:05 ` Michael Buesch
2007-12-14 19:25 ` Ray Lee
2007-12-14 19:38 ` Michael Buesch
2007-12-14 19:55 ` Ray Lee
2007-12-14 20:13 ` Michael Buesch
2007-12-14 21:39 ` Ray Lee
2007-12-14 12:59 ` Simon Holm Thøgersen
2007-12-14 13:04 ` Michael Buesch
2007-12-15 0:51 ` Rafael J. Wysocki
2007-12-15 0:58 ` Larry Finger
2007-12-15 1:59 ` mvtodevnull
2007-12-15 2:27 ` Larry Finger
2007-12-15 4:37 ` mvtodevnull
2007-12-15 4:39 ` mvtodevnull
2007-12-15 7:18 ` Larry Finger
2007-12-15 12:38 ` mvtodevnull
2007-12-17 2:17 ` mvtodevnull
2007-12-17 6:52 ` Larry Finger
2007-12-17 7:17 ` mvtodevnull
2007-12-17 9:49 ` Michael Buesch
2007-12-17 10:15 ` mvtodevnull
2007-12-17 10:35 ` mvtodevnull
2007-12-17 22:04 ` mvtodevnull
2007-12-17 22:45 ` Michael Buesch
2007-12-17 23:12 ` mvtodevnull
2007-12-17 23:18 ` Michael Buesch
2007-12-17 23:27 ` mvtodevnull
2007-12-18 1:16 ` Larry Finger
2007-12-18 2:34 ` mvtodevnull
2007-12-19 2:12 ` mvtodevnull
2007-12-19 8:11 ` Larry Finger
[not found] ` <200712191046.54147.mb@bu3sch.de>
2007-12-19 13:48 ` Ingo Molnar
2007-12-15 10:57 ` Michael Buesch
2007-12-15 23:18 ` Rafael J. Wysocki
2007-12-15 23:27 ` Michael Buesch
2007-12-16 2:30 ` Larry Finger
2007-12-16 11:13 ` Michael Buesch
2007-12-16 13:56 ` Johannes Berg
2007-12-16 14:35 ` Rafael J. Wysocki
2007-12-16 14:18 ` Johannes Berg
2007-12-14 2:09 ` Larry Finger
2007-12-14 10:56 ` Ingo Molnar
2007-12-14 14:14 ` John W. Linville
2007-12-14 14:40 ` stefano.brivio
2007-12-14 16:27 ` Ray Lee
2007-12-14 16:45 ` Ray Lee
2007-12-14 16:59 ` Michael Buesch
2007-12-14 17:22 ` Ray Lee
2007-12-14 17:59 ` Ingo Molnar
2007-12-14 18:38 ` Michael Buesch
2007-12-15 1:25 ` Rafael J. Wysocki
2007-12-15 21:43 ` John W. Linville
2007-12-15 23:26 ` Rafael J. Wysocki
2007-12-16 3:20 ` Stefan Lippers-Hollmann
2007-12-16 9:22 ` Ingo Molnar
2007-12-16 11:23 ` Michael Buesch
2007-12-13 1:20 ` Daniel Walker
2007-12-13 10:13 ` Ingo Molnar
2007-12-13 11:11 ` Michael Buesch
2007-12-13 14:23 ` John W. Linville
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=47688A5A.1090902@am.sony.com \
--to=geoffrey.levand@am.sony.com \
--cc=akpm@linux-foundation.org \
--cc=dwalker@mvista.com \
--cc=kjwinchester@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@bohmer.net \
--cc=linuxppc-dev@ozlabs.org \
--cc=matthias.kaehlcke@gmail.com \
--cc=mingo@elte.hu \
/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