linux-kernel - Re: [PATCH] ARM: imx: Drop imx_anatop

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: <VI1PR04MB70238747C8EDB6C9A67700A3EEA50@VI1PR04MB7023.eurprd04.prod.outlook.com>
Date:   Thu, 22 Aug 2019 23:26:47 +0000
From:   Leonard Crestez <leonard.crestez@....com>
To:     Andrey Smirnov <andrew.smirnov@...il.com>
CC:     Shawn Guo <shawnguo@...nel.org>,
        Fabio Estevam <festevam@...il.com>,
        Chris Healy <cphealy@...il.com>,
        Peter Chen <peter.chen@....com>,
        dl-linux-imx <linux-imx@....com>,
        "linux-arm-kernel@...ts.infradead.org" 
        <linux-arm-kernel@...ts.infradead.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Aisheng Dong <aisheng.dong@....com>
Subject: Re: [PATCH] ARM: imx: Drop imx_anatop_init()

On 22.08.2019 23:06, Andrey Smirnov wrote:
>> On 31.07.2019 21:01, Andrey Smirnov wrote:
>>> With commit b5bbe2235361 ("usb: phy: mxs: Disable external charger
>>> detect in mxs_phy_hw_init()") in tree all of the necessary charger
>>> setup is done by the USB PHY driver which covers all of the affected
>>> i.MX6 SoCs.
>>>
>>> NOTE: Imx_anatop_init() was also called for i.MX7D, but looking at its
>>> datasheet it appears to have a different USB PHY IP block, so
>>> executing i.MX6 charger disable configuration seems unnecessary.
>>>
>>> -void __init imx_anatop_init(void)
>>> -{
>>> -     anatop = syscon_regmap_lookup_by_compatible("fsl,imx6q-anatop");
>>> -     if (IS_ERR(anatop)) {
>>> -             pr_err("%s: failed to find imx6q-anatop regmap!\n", __func__);
>>> -             return;
>>> -     }
>>
>> This patch breaks suspend on imx6 in linux-next because the "anatop"
>> regmap is no longer initialized. This was found via bisect but
>> no_console_suspend prints a helpful stack anyway:
>>
>> (regmap_read) from [<c01226e4>] (imx_anatop_enable_weak2p5+0x28/0x70)
>> (imx_anatop_enable_weak2p5) from [<c0122744>]
>> (imx_anatop_pre_suspend+0x18/0x64)
>> (imx_anatop_pre_suspend) from [<c0124434>] (imx6q_pm_enter+0x60/0x16c)
>> (imx6q_pm_enter) from [<c018c8a4>] (suspend_devices_and_enter+0x7d4/0xcbc)
>> (suspend_devices_and_enter) from [<c018d544>] (pm_suspend+0x7b8/0x904)
>> (pm_suspend) from [<c018b1b4>] (state_store+0x68/0xc8)
>>
> 
> My bad, completely missed that fact that anatop was a global variable
> in  imx_anatop_init(). Sorry about that.
> 
>> Minimal fix looks like this:
>>
>> --- arch/arm/mach-imx/anatop.c
>> +++ arch/arm/mach-imx/anatop.c
>> @@ -111,6 +111,12 @@ void __init imx_init_revision_from_anatop(void)
>>            digprog = readl_relaxed(anatop_base + offset);
>>            iounmap(anatop_base);
>>
>> +       anatop = syscon_regmap_lookup_by_compatible("fsl,imx6q-anatop");
>> +       if (IS_ERR(anatop)) {
>> +               pr_err("failed to find imx6q-anatop regmap!\n");
>> +               return;
>> +       }
>>
>> Since all SOCs that called imx_anatop_init also call
>> imx_init_revision_from_anatop this might be an acceptable solution,
>> unless there is some limitation preventing early regmap lookup.
>>
> 
> Would making every function that uses anatop explicitly request it via
> syscon_regmap_lookup_by_compatible("fsl,imx6q-anatop") be too much of
> a code duplication? This way we won't need to worry if
> imx_init_revision_from_anatop() was called before any of them are
> used.

It's only used from pre_suspend and post_suspend, everything else in 
arch/arm/mach-imx/anatop.c is static. Doing a lookup every time would be 
silly, it's fine to let this be global.

I was wondering if maybe imx_init_revision could somehow end up getting 
called before syscon/regmap core init.

--
Regards,
Leonard