From: Nam Cao <namcao@linutronix.de>
To: Yunseong Kim <ysk@kzalloc.com>
Cc: gregkh@linuxfoundation.org, stern@rowland.harvard.edu,
linux-usb@vger.kernel.org, "Thomas Gleixner" <tglx@linutronix.de>,
"Sebastian Andrzej Siewior" <bigeasy@linutronix.de>,
"Clark Williams" <clrkwllms@kernel.org>,
"Steven Rostedt" <rostedt@goodmis.org>,
"Marcello Sylvester Bauer" <sylv@sylv.io>,
"Krzysztof Kozlowski" <krzysztof.kozlowski@linaro.org>,
"Uwe Kleine-König" <u.kleine-koenig@baylibre.com>,
"Al Viro" <viro@zeniv.linux.org.uk>,
andreyknvl@gmail.com, "Austin Kim" <austindh.kim@gmail.com>,
linux-rt-users@vger.kernel.org, linux-kernel@vger.kernel.org,
syzkaller@googlegroups.com
Subject: Re: [BUG] usb: gadget: dummy_hcd: Sleeping function called from invalid context in dummy_dequeue on PREEMPT_RT
Date: Sat, 16 Aug 2025 08:59:33 +0200 [thread overview]
Message-ID: <20250816065933.EPwBJ0Sd@linutronix.de> (raw)
In-Reply-To: <5b337389-73b9-4ee4-a83e-7e82bf5af87a@kzalloc.com>
On Sat, Aug 16, 2025 at 11:38:14AM +0900, Yunseong Kim wrote:
> While testing a PREEMPT_RT enabled kernel (based on v6.17.0-rc1),
> I encountered a "BUG: sleeping function called from invalid context" error
> originating from the dummy_dequeue function in the dummy USB driver.
...
> The pattern of manually disabling IRQs and then taking a spinlock
> local_irq_save() + spin_lock() is unsafe on PREEMPT_RT, the current code
> structure keeps IRQs disabled even after spin_unlock(&dum->lock) while
> calling usb_gadget_giveback_request(). This extended atomic context can
> also be problematic if the completion handler attempts to acquire another
> sleepable lock.
I don't know the USB subsystem well, but the comments above struct
usb_request says:
* @complete: Function called when request completes, so this request and
* its buffer may be re-used. The function will always be called with
* interrupts disabled, and it must not sleep.
Therefore it shouldn't be a concern that "completion handler attempts to
acquire another sleepable lock".
> I request a review and correction of this locking mechanism to ensure
> stability on PREEMPT_RT configurations. Kernel config, full logs, and
> reproduction steps can be provided on request.
This was introduced by b4dbda1a22d2 ("USB: dummy-hcd: disable interrupts
during req->complete") which split the spin_lock_irqsave() into
local_irq_save() and spin_lock().
The untested patch below should help?
Enabling interrupt (spin_unlock_irqrestore) and then immediately disabling
interrupt (local_irq_save) is not the nicest thing. But then I don't see
how to avoid that while being non-hacky and human-readable.
Nam
diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c
index 21dbfb0b3bac..a4653c919664 100644
--- a/drivers/usb/gadget/udc/dummy_hcd.c
+++ b/drivers/usb/gadget/udc/dummy_hcd.c
@@ -765,8 +765,7 @@ static int dummy_dequeue(struct usb_ep *_ep, struct usb_request *_req)
if (!dum->driver)
return -ESHUTDOWN;
- local_irq_save(flags);
- spin_lock(&dum->lock);
+ spin_lock_irqsave(&dum->lock, flags);
list_for_each_entry(iter, &ep->queue, queue) {
if (&iter->req != _req)
continue;
@@ -776,15 +775,16 @@ static int dummy_dequeue(struct usb_ep *_ep, struct usb_request *_req)
retval = 0;
break;
}
- spin_unlock(&dum->lock);
+ spin_unlock_irqrestore(&dum->lock, flags);
if (retval == 0) {
dev_dbg(udc_dev(dum),
"dequeued req %p from %s, len %d buf %p\n",
req, _ep->name, _req->length, _req->buf);
+ local_irq_save(flags);
usb_gadget_giveback_request(_ep, _req);
+ local_irq_restore(flags);
}
- local_irq_restore(flags);
return retval;
}
next prev parent reply other threads:[~2025-08-16 6:59 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-16 2:38 [BUG] usb: gadget: dummy_hcd: Sleeping function called from invalid context in dummy_dequeue on PREEMPT_RT Yunseong Kim
2025-08-16 6:59 ` Nam Cao [this message]
2025-08-16 11:41 ` Yunseong Kim
2025-08-19 10:21 ` Sebastian Andrzej Siewior
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=20250816065933.EPwBJ0Sd@linutronix.de \
--to=namcao@linutronix.de \
--cc=andreyknvl@gmail.com \
--cc=austindh.kim@gmail.com \
--cc=bigeasy@linutronix.de \
--cc=clrkwllms@kernel.org \
--cc=gregkh@linuxfoundation.org \
--cc=krzysztof.kozlowski@linaro.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rt-users@vger.kernel.org \
--cc=linux-usb@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=stern@rowland.harvard.edu \
--cc=sylv@sylv.io \
--cc=syzkaller@googlegroups.com \
--cc=tglx@linutronix.de \
--cc=u.kleine-koenig@baylibre.com \
--cc=viro@zeniv.linux.org.uk \
--cc=ysk@kzalloc.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).