linux-kernel - Re: [PATCH] HID: intel-ish-hid: Replace PCI_DEV_FLAGS_NO_D3 with pci_save

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: <7E88D4A8-8056-4E12-8B2C-27307A7C5E7D@canonical.com>
Date:   Thu, 21 May 2020 12:43:15 +0800
From:   Kai-Heng Feng <kai.heng.feng@...onical.com>
To:     Srinivas Pandruvada <srinivas.pandruvada@...ux.intel.com>
Cc:     "Rafael J. Wysocki" <rafael.j.wysocki@...el.com>,
        Jiri Kosina <jikos@...nel.org>,
        Benjamin Tissoires <benjamin.tissoires@...hat.com>,
        Zhang Lixu <lixu.zhang@...el.com>, Even Xu <even.xu@...el.com>,
        Alexios Zavras <alexios.zavras@...el.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Song Hongyan <hongyan.song@...el.com>,
        "open list:INTEL INTEGRATED SENSOR HUB DRIVER" 
        <linux-input@...r.kernel.org>,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] HID: intel-ish-hid: Replace PCI_DEV_FLAGS_NO_D3 with
 pci_save_state

Hi Srinivas,

> On May 9, 2020, at 01:45, Srinivas Pandruvada <srinivas.pandruvada@...ux.intel.com> wrote:
> 
> On Tue, 2020-05-05 at 21:17 +0800, Kai-Heng Feng wrote:
>> PCI_DEV_FLAGS_NO_D3 should not be used outside of PCI core.
>> 
>> Instead, we can use pci_save_state() to hint PCI core that the device
>> should stay at D0 during suspend.
> 
> Your changes are doing more than just changing the flag. Can you
> explain more about the other changes?

By using pci_save_state(), in addition to keep itself stay at D0, the parent bridge will also stay at D0.
So it's a better approach to achieve the same thing.

> Also make sure that you test on both platforms which has regular S3 and
> S0ix (modern standby system).

Actually I don't have any physical hardware to test the patch, I found the issue when I search for D3 quirks through the source code.

Can you guys do a quick smoketest for this patch?

Kai-Heng

> 
> Thanks,
> Srinivas
> 
> 
>> 
>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@...onical.com>
>> ---
>> drivers/hid/intel-ish-hid/ipc/pci-ish.c | 15 ++++++++++-----
>> 1 file changed, 10 insertions(+), 5 deletions(-)
>> 
>> diff --git a/drivers/hid/intel-ish-hid/ipc/pci-ish.c
>> b/drivers/hid/intel-ish-hid/ipc/pci-ish.c
>> index f491d8b4e24c..ab588b9c8d09 100644
>> --- a/drivers/hid/intel-ish-hid/ipc/pci-ish.c
>> +++ b/drivers/hid/intel-ish-hid/ipc/pci-ish.c
>> @@ -106,6 +106,11 @@ static inline bool ish_should_enter_d0i3(struct
>> pci_dev *pdev)
>> 	return !pm_suspend_via_firmware() || pdev->device ==
>> CHV_DEVICE_ID;
>> }
>> 
>> +static inline bool ish_should_leave_d0i3(struct pci_dev *pdev)
>> +{
>> +	return !pm_resume_via_firmware() || pdev->device ==
>> CHV_DEVICE_ID;
>> +}
>> +
>> /**
>>  * ish_probe() - PCI driver probe callback
>>  * @pdev:	pci device
>> @@ -215,9 +220,7 @@ static void __maybe_unused
>> ish_resume_handler(struct work_struct *work)
>> 	struct ishtp_device *dev = pci_get_drvdata(pdev);
>> 	int ret;
>> 
>> -	/* Check the NO_D3 flag to distinguish the resume paths */
>> -	if (pdev->dev_flags & PCI_DEV_FLAGS_NO_D3) {
>> -		pdev->dev_flags &= ~PCI_DEV_FLAGS_NO_D3;
>> +	if (ish_should_leave_d0i3(pdev) && !dev->suspend_flag) {
>> 		disable_irq_wake(pdev->irq);
>> 
>> 		ishtp_send_resume(dev);
>> @@ -281,8 +284,10 @@ static int __maybe_unused ish_suspend(struct
>> device *device)
>> 			 */
>> 			ish_disable_dma(dev);
>> 		} else {
>> -			/* Set the NO_D3 flag, the ISH would enter D0i3
>> */
>> -			pdev->dev_flags |= PCI_DEV_FLAGS_NO_D3;
>> +			/* Save state so PCI core will keep the device
>> at D0,
>> +			 * the ISH would enter D0i3
>> +			 */
>> +			pci_save_state(pdev);
>> 
> Did you test on some C
> 
> 
>> 			enable_irq_wake(pdev->irq);
>> 		}