linux-kernel - Re: [External] Re: [PATCH v2 1/2] perf/dwc_pcie: fix the incorrect reference count

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAEEQ3wm6vw9AvEtWOaPqLvR3x4vs=KMACN9OQGJ2eAOzuv2GhQ@mail.gmail.com>
Date: Sat, 8 Feb 2025 10:57:13 +0800
From: yunhui cui <cuiyunhui@...edance.com>
To: Shuai Xue <xueshuai@...ux.alibaba.com>
Cc: renyu.zj@...ux.alibaba.com, will@...nel.org, mark.rutland@....com, 
	linux-arm-kernel@...ts.infradead.org, linux-perf-users@...r.kernel.org, 
	linux-kernel@...r.kernel.org, Bjorn Helgaas <bhelgaas@...gle.com>
Subject: Re: [External] Re: [PATCH v2 1/2] perf/dwc_pcie: fix the incorrect
 reference count

Hi Shuai,


On Fri, Feb 7, 2025 at 10:20 AM Shuai Xue <xueshuai@...ux.alibaba.com> wrote:
>
>
>
> 在 2025/2/5 11:14, Yunhui Cui 写道:
> > for_each_pci_dev increments pdev refcount. Call pci_dev_put after access
> > to decrement it.
> >
> > Signed-off-by: Yunhui Cui <cuiyunhui@...edance.com>
> > ---
> >   drivers/perf/dwc_pcie_pmu.c | 42 ++++++++++++++++++++++++++-----------
> >   1 file changed, 30 insertions(+), 12 deletions(-)
> >
> > diff --git a/drivers/perf/dwc_pcie_pmu.c b/drivers/perf/dwc_pcie_pmu.c
> > index cccecae9823f..4ac53167d7ab 100644
> > --- a/drivers/perf/dwc_pcie_pmu.c
> > +++ b/drivers/perf/dwc_pcie_pmu.c
> > @@ -553,6 +553,7 @@ static u16 dwc_pcie_des_cap(struct pci_dev *pdev)
> >
> >   static void dwc_pcie_unregister_dev(struct dwc_pcie_dev_info *dev_info)
> >   {
> > +     pci_dev_put(dev_info->pdev);
> >       platform_device_unregister(dev_info->plat_dev);
> >       list_del(&dev_info->dev_node);
> >       kfree(dev_info);
> > @@ -572,8 +573,10 @@ static int dwc_pcie_register_dev(struct pci_dev *pdev)
> >               return PTR_ERR(plat_dev);
> >
> >       dev_info = kzalloc(sizeof(*dev_info), GFP_KERNEL);
> > -     if (!dev_info)
> > +     if (!dev_info) {
> > +             platform_device_unregister(plat_dev);
> >               return -ENOMEM;
> > +     }
> >
> >       /* Cache platform device to handle pci device hotplug */
> >       dev_info->plat_dev = plat_dev;
> > @@ -594,8 +597,11 @@ static int dwc_pcie_pmu_notifier(struct notifier_block *nb,
> >       case BUS_NOTIFY_ADD_DEVICE:
> >               if (!dwc_pcie_des_cap(pdev))
> >                       return NOTIFY_DONE;
> > -             if (dwc_pcie_register_dev(pdev))
> > +             pci_dev_get(pdev);
> > +             if (dwc_pcie_register_dev(pdev)) {
> > +                     pci_dev_put(pdev);
> >                       return NOTIFY_BAD;
> > +             }
> >               break;
> >       case BUS_NOTIFY_DEL_DEVICE:
> >               dev_info = dwc_pcie_find_dev_info(pdev);
> > @@ -730,20 +736,29 @@ static struct platform_driver dwc_pcie_pmu_driver = {
> >       .driver = {.name = "dwc_pcie_pmu",},
> >   };
> >
> > +static void dwc_pcie_cleanup_devices(void)
> > +{
> > +     struct dwc_pcie_dev_info *dev_info, *tmp;
> > +
> > +     list_for_each_entry_safe(dev_info, tmp, &dwc_pcie_dev_info_head, dev_node) {
> > +             dwc_pcie_unregister_dev(dev_info);
> > +     }
> > +}
> > +
> >   static int __init dwc_pcie_pmu_init(void)
> >   {
> >       struct pci_dev *pdev = NULL;
> >       int ret;
> >
> >       for_each_pci_dev(pdev) {
> > -             if (!dwc_pcie_des_cap(pdev))
> > +             if (!dwc_pcie_des_cap(pdev)) {
> > +                     pci_dev_put(pdev);
>
> If the pdev has no RAS cap, we do not need to put the device.
>
> > The reference count for @from is
> > always decremented if it is not %NULL.
>
> Please see comments from pci_get_device() for more details.
Oh, yes. If that's the case, I'll update to v3.


> Shuai
>

Thanks，
Yunhui