From: Hans de Goede <hdegoede@redhat.com>
To: lizhe <sensor1010@163.com>,
bhelgaas@google.com, lukas@wunner.de, ameynarkhede03@gmail.com,
naveennaidu479@gmail.com
Cc: linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] PCI:pciehp: Replace request_threaded_irq with devm_request_threaded_irq
Date: Sun, 6 Feb 2022 15:34:56 +0100 [thread overview]
Message-ID: <4b9d7ddc-9e91-5fbf-da2a-f2bd4b52beec@redhat.com> (raw)
In-Reply-To: <20220206140705.10705-1-sensor1010@163.com>
Hi Lihze,
On 2/6/22 15:07, lizhe wrote:
> Memory allocated with this function is automatically
> freed on driver detach
>
> Signed-off-by: lizhe <sensor1010@163.com>
You must use your real name (first-name + last-name) as author,
as well as in the Signed-off-by line, see:
https://www.kernel.org/doc/html/latest/process/submitting-patches.html#sign-your-work-the-developer-s-certificate-of-origin
> ---
> drivers/pci/hotplug/pciehp_hpc.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/pci/hotplug/pciehp_hpc.c b/drivers/pci/hotplug/pciehp_hpc.c
> index 1c1ebf3dad43..aca59c6fdcbc 100644
> --- a/drivers/pci/hotplug/pciehp_hpc.c
> +++ b/drivers/pci/hotplug/pciehp_hpc.c
> @@ -66,7 +66,7 @@ static inline int pciehp_request_irq(struct controller *ctrl)
> }
>
> /* Installs the interrupt handler */
> - retval = request_threaded_irq(irq, pciehp_isr, pciehp_ist,
> + retval = devm_request_threaded_irq(irq, pciehp_isr, pciehp_ist,
> IRQF_SHARED, "pciehp", ctrl);
> if (retval)
> ctrl_err(ctrl, "Cannot get irq %d for the hotplug controller\n",
> @@ -78,8 +78,6 @@ static inline void pciehp_free_irq(struct controller *ctrl)
> {
> if (pciehp_poll_mode)
> kthread_stop(ctrl->poll_thread);
> - else
> - free_irq(ctrl->pcie->irq, ctrl);
> }
>
> static int pcie_poll_cmd(struct controller *ctrl, int timeout)
>
You cannot just go and change a single allocation into a devm managed
resource, esp. not a request_irq call.
Changing this into a devm_allocation means that the irq now will not be
free-ed until after pciehp_remove() has completed and that function calls:
pciehp_release_ctrl(ctrl); which releases the memory the ctrl pointer points
to and that same memory / pointer is used by pciehp_isr.
So after your patch it is possible for the IRQ to trigger after
pciehp_release_ctrl(ctrl) has free-ed the memory (and before the devm
framework calls free_irq() disabling the IRQ), causing pciehp_isr
to reference free-ed memory, leading to a crash.
Since this patch introduces a bug we cannot accept it, nack.
Regards,
Hans
prev parent reply other threads:[~2022-02-06 14:37 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-06 14:07 lizhe
2022-02-06 14:34 ` Hans de Goede [this message]
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=4b9d7ddc-9e91-5fbf-da2a-f2bd4b52beec@redhat.com \
--to=hdegoede@redhat.com \
--cc=ameynarkhede03@gmail.com \
--cc=bhelgaas@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=lukas@wunner.de \
--cc=naveennaidu479@gmail.com \
--cc=sensor1010@163.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
all inboxes | Powered by JetHome®