commit 0583bc776ca5b5a3f5752869fc31cf7322df2b35 upstream.
dwc2_hc_n_intr() writes back INTMASK as read but evaluates it
with intmask applied. In stress testing this causes spurious
interrupts like this:
[Mon Aug 14 10:51:07 2023] dwc2 3f980000.usb: dwc2_hc_chhltd_intr_dma: Channel 7 - ChHltd set, but reason is unknown
[Mon Aug 14 10:51:07 2023] dwc2 3f980000.usb: hcint 0x00000002, intsts 0x04600001
[Mon Aug 14 10:51:08 2023] dwc2 3f980000.usb: dwc2_hc_chhltd_intr_dma: Channel 0 - ChHltd set, but reason is unknown
[Mon Aug 14 10:51:08 2023] dwc2 3f980000.usb: hcint 0x00000002, intsts 0x04600001
[Mon Aug 14 10:51:08 2023] dwc2 3f980000.usb: dwc2_hc_chhltd_intr_dma: Channel 4 - ChHltd set, but reason is unknown
[Mon Aug 14 10:51:08 2023] dwc2 3f980000.usb: hcint 0x00000002, intsts 0x04600001
[Mon Aug 14 10:51:08 2023] dwc2 3f980000.usb: dwc2_update_urb_state_abn(): trimming xfer length
Applying INTMASK prevents this. The issue exists in all versions of the
driver.
Signed-off-by: Oliver Neukum <oneukum@suse.com>
Tested-by: Ivan Ivanov <ivan.ivanov@suse.com>
Tested-by: Andrea della Porta <andrea.porta@suse.com>
Link: https://lore.kernel.org/r/20231115144514.15248-1-oneukum@suse.com
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit ef307bc6ef04e8c1ea843231db58e3afaafa9fa6 ]
In _dwc2_hcd_urb_enqueue(), "urb->hcpriv = NULL" is executed without
holding the lock "hsotg->lock". In _dwc2_hcd_urb_dequeue():
spin_lock_irqsave(&hsotg->lock, flags);
...
if (!urb->hcpriv) {
dev_dbg(hsotg->dev, "## urb->hcpriv is NULL ##\n");
goto out;
}
rc = dwc2_hcd_urb_dequeue(hsotg, urb->hcpriv); // Use urb->hcpriv
...
out:
spin_unlock_irqrestore(&hsotg->lock, flags);
When _dwc2_hcd_urb_enqueue() and _dwc2_hcd_urb_dequeue() are
concurrently executed, the NULL check of "urb->hcpriv" can be executed
before "urb->hcpriv = NULL". After urb->hcpriv is NULL, it can be used
in the function call to dwc2_hcd_urb_dequeue(), which can cause a NULL
pointer dereference.
This possible bug is found by an experimental static analysis tool
developed by myself. This tool analyzes the locking APIs to extract
function pairs that can be concurrently executed, and then analyzes the
instructions in the paired functions to identify possible concurrency
bugs including data races and atomicity violations. The above possible
bug is reported, when my tool analyzes the source code of Linux 6.5.
To fix this possible bug, "urb->hcpriv = NULL" should be executed with
holding the lock "hsotg->lock". After using this patch, my tool never
reports the possible bug, with the kernelconfiguration allyesconfig for
x86_64. Because I have no associated hardware, I cannot test the patch
in runtime testing, and just verify it according to the code logic.
Fixes: 33ad261aa62b ("usb: dwc2: host: spinlock urb_enqueue")
Signed-off-by: Jia-Ju Bai <baijiaju@buaa.edu.cn>
Link: https://lore.kernel.org/r/20230926024404.832096-1-baijiaju@buaa.edu.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
