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-next>] [day] [month] [year] [list]
Message-ID: <da4c75ba87c3476694361ee3bc333401@realtek.com>
Date:   Fri, 3 Nov 2023 10:54:26 +0000
From:   Stanley Chang[昌育德] 
        <stanley_chang@...ltek.com>
To:     Christophe JAILLET <christophe.jaillet@...adoo.fr>
CC:     "Thinh.Nguyen@...opsys.com" <Thinh.Nguyen@...opsys.com>,
        "conor+dt@...nel.org" <conor+dt@...nel.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
        "gregkh@...uxfoundation.org" <gregkh@...uxfoundation.org>,
        "krzysztof.kozlowski+dt@...aro.org" 
        <krzysztof.kozlowski+dt@...aro.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-usb@...r.kernel.org" <linux-usb@...r.kernel.org>,
        "robh+dt@...nel.org" <robh+dt@...nel.org>
Subject: RE: RE: [PATCH v6 1/2] usb: dwc3: add Realtek DHC RTD SoC dwc3 glue layer driver

Hi Christophe,

> Le 01/11/2023 à 07:27, Stanley Chang[昌育德] a écrit :
> > Hi CJ,
> >
> > I think these functions are not needed in remove function.
> >
> > In dwc3_rtk_probe_dwc3_core,
> > I have used
> > dwc3_node = of_get_compatible_child(node, "snps,dwc3"); and dwc3_pdev
> > = of_find_device_by_node(dwc3_node);
> >
> > So, I call these put functions.
> > platform_device_put(dwc3_pdev);
> > of_node_put(dwc3_node);
> 
> Yes, but you call it only in the error handling path of the function.
> 
> I wonder if they should also be called in the remove function in order to
> decrement the ref-counted reference.
> 
> 
> Same in __get_dwc3_maximum_speed(), the reference taken by:
>     dwc3_np = of_get_compatible_child(np, "snps,dwc3"); is never released.
> 
> 
> See the comment at [1] to see what I mean.
> 
> 
> [1]: https://elixir.bootlin.com/linux/v6.6/source/drivers/of/base.c#L681

You are right!
For dwc3_pdev, dwc3_np or dwc3_node, I should add of_node_put to release them when the function exits. 
https://elixir.bootlin.com/linux/v6.5.10/source/drivers/usb/dwc3/dwc3-meson-g12a.c#L567

I will add a patch to fix this.

For example,
diff --git a/drivers/usb/dwc3/dwc3-rtk.c b/drivers/usb/dwc3/dwc3-rtk.c
index 590028e8fdcb..9d6f2a8bd6ce 100644
--- a/drivers/usb/dwc3/dwc3-rtk.c
+++ b/drivers/usb/dwc3/dwc3-rtk.c
@@ -187,6 +187,7 @@ static enum usb_device_speed __get_dwc3_maximum_speed(struct device_node *np)

        ret = match_string(speed_names, ARRAY_SIZE(speed_names), maximum_speed);

+       of_node_put(dwc3_np);
        return (ret < 0) ? USB_SPEED_UNKNOWN : ret;
 }

@@ -339,6 +340,8 @@ static int dwc3_rtk_probe_dwc3_core(struct dwc3_rtk *rtk)

        switch_usb2_role(rtk, rtk->cur_role);

+       platform_device_put(dwc3_pdev);
+       of_node_put(dwc3_node);
        return 0;

 err_pdev_put:


Thanks,
Stanley

> CJ
> >
> > Thanks,
> > Stanley
> >
> >> Hi,
> >>
> >> Is something like
> >>          platform_device_put(dwc3_pdev);
> >>          of_node_put(dwc3_node);
> >> needed in the remove function?
> >>
> >> (as done in the error handling path of dwc3_rtk_probe_dwc3_core())
> >>
> >> Or should it be added at the end of dwc3_rtk_probe_dwc3_core() if the
> >> reference are nor needed anymore when we leave the function?
> >>
> >> CJ
> >>
> >>> +     of_platform_depopulate(rtk->dev); }
> >>> +
> >>
> >> ...
> >

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ