| 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
| ||
|
Date: Wed, 6 Feb 2019 20:53:33 +0530
From: Alim Akhtar <alim.akhtar@...sung.com>
To: Bjorn Andersson <bjorn.andersson@...aro.org>
Cc: Marc Gonzalez <marc.w.gonzalez@...e.fr>,
MSM <linux-arm-msm@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>,
Jeffrey Hugo <jhugo@...eaurora.org>,
Andy Gross <andy.gross@...aro.org>,
David Brown <david.brown@...aro.org>,
Evan Green <evgreen@...omium.org>,
Douglas Anderson <dianders@...omium.org>,
Avri Altman <avri.altman@....com>,
Pedro Sousa <pedrom.sousa@...opsys.com>,
Subhash Jadavani <subhashj@...eaurora.org>,
Bart Van Assche <bvanassche@....org>,
SCSI <linux-scsi@...r.kernel.org>
Subject: Re: [PATCH v3 5/5] Revert
"scsi: ufs: disable vccq if it's not needed by UFS device"
On 06/02/19 1:16 AM, Bjorn Andersson wrote:
> On Tue 05 Feb 02:52 PST 2019, Alim Akhtar wrote:
>
>> Hi Bjorn,
>>
>> On 05/02/19 11:57 AM, Bjorn Andersson wrote:
>>> On Mon 04 Feb 20:58 PST 2019, Alim Akhtar wrote:
>>>
>>>> Hi Marc,
>>>>
>>>> On 04/02/19 11:12 PM, Marc Gonzalez wrote:
>>>>> This reverts commit 60f0187031c05e04cbadffb62f557d0ff3564490.
>>>>>
>>>>> Calling ufshcd_set_vccq_rail_unused hangs my system.
>>>>> It seems vccq is not *not* needed.
>>>>>
>>>>> Signed-off-by: Marc Gonzalez <marc.w.gonzalez@...e.fr>
>>>>> ---
>>>>
>>>> AFAIK Samsung and Toshiba UFS devices does not use VCCQ (this pin is
>>>> either floating or connected to Ground, at least on the devices that I
>>>> have worked on).
>>>
>>> But why does such system define a vccq-supply? If the system doesn't
>>> have a regulator connected to VCCQ, then the UFS driver shouldn't be
>>> told that there is one. And if VCCQ is optional the UFS driver should
>>> support the fact that this regulator might not be supplied (i.e. call
>>> regulator_get_optional() and handle the error indicating that the supply
>>> isn't specified).
>>>
>> As per JESD220C, chapter 6.1, it does says "VCCQ - Supply voltage used
>> typically for the memory controller and optionally for the PHY
>> interface, the memory IO, and any other internal very low voltage block"
>> And we have VCCQ2 - which serve the pretty much same purpose. The
>> voltage range for VCCQ and VCCQ2 are different, VCCQ has a lower voltage
>> suitable to some low voltage block inside UFS device. I think this is
>> design consideration which allow some vendor to use one less physical
>> pin may be. And also depends on the voltage requirements of some of the
>> internal circuit.
>>
>
> This looks to me that you are required to have a VCCQ. But you said that
> you do not have a regulator supplying VCCQ on your board, and if that
> really is the case then you should not specify vccq-supply.
>
> The patch Marc is reverting states that for devices that does not have
> VCCQ connected, some unrelated regulator should be assigned to the UFS
> driver so that it can turn it off.
>
>
> If you have a regulator connected to VCCQ then it should go in
> vccq-supply, if you have a regulator connected to VCCQ2 the is should go
> in vccq2-supply. If you don't have these pins connected then there
> shouldn't be any regulators specified here!
>
Yes, that's correct, it should be like what you are suggesting, DT
entries should match the board schematic.
> Regards,
> Bjorn
>
>>
>>> Regards,
>>> Bjorn
>>>
>>>> You said your system hanged, I believe you have set UFS_DEVICE_NO_VCCQ
>>>> quirks, in that case VCCQ regulator should having been disabled.
>>>> So you mean your system hanged because vccq regulator got disabled?
>>>>
>>>>> drivers/scsi/ufs/ufs.h | 1 -
>>>>> drivers/scsi/ufs/ufshcd.c | 59 +++------------------------------------
>>>>> 2 files changed, 4 insertions(+), 56 deletions(-)
>>>>>
>>>>> diff --git a/drivers/scsi/ufs/ufs.h b/drivers/scsi/ufs/ufs.h
>>>>> index dd65fea07687..7da7318eb6a6 100644
>>>>> --- a/drivers/scsi/ufs/ufs.h
>>>>> +++ b/drivers/scsi/ufs/ufs.h
>>>>> @@ -514,7 +514,6 @@ struct ufs_vreg {
>>>>> struct regulator *reg;
>>>>> const char *name;
>>>>> bool enabled;
>>>>> - bool unused;
>>>>> int min_uV;
>>>>> int max_uV;
>>>>> int min_uA;
>>>>> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
>>>>> index 9ba7671b84f8..8b9a01073d62 100644
>>>>> --- a/drivers/scsi/ufs/ufshcd.c
>>>>> +++ b/drivers/scsi/ufs/ufshcd.c
>>>>> @@ -245,7 +245,6 @@ static int ufshcd_probe_hba(struct ufs_hba *hba);
>>>>> static int __ufshcd_setup_clocks(struct ufs_hba *hba, bool on,
>>>>> bool skip_ref_clk);
>>>>> static int ufshcd_setup_clocks(struct ufs_hba *hba, bool on);
>>>>> -static int ufshcd_set_vccq_rail_unused(struct ufs_hba *hba, bool unused);
>>>>> static int ufshcd_uic_hibern8_exit(struct ufs_hba *hba);
>>>>> static int ufshcd_uic_hibern8_enter(struct ufs_hba *hba);
>>>>> static inline void ufshcd_add_delay_before_dme_cmd(struct ufs_hba *hba);
>>>>> @@ -6819,11 +6818,6 @@ static int ufshcd_probe_hba(struct ufs_hba *hba)
>>>>> ufs_fixup_device_setup(hba, &card);
>>>>> ufshcd_tune_unipro_params(hba);
>>>>>
>>>>> - ret = ufshcd_set_vccq_rail_unused(hba,
>>>>> - (hba->dev_quirks & UFS_DEVICE_NO_VCCQ) ? true : false);
>>>>> - if (ret)
>>>>> - goto out;
>>>>> -
>>>>> /* UFS device is also active now */
>>>>> ufshcd_set_ufs_dev_active(hba);
>>>>> ufshcd_force_reset_auto_bkops(hba);
>>>>> @@ -7007,24 +7001,13 @@ static int ufshcd_config_vreg_load(struct device *dev, struct ufs_vreg *vreg,
>>>>> static inline int ufshcd_config_vreg_lpm(struct ufs_hba *hba,
>>>>> struct ufs_vreg *vreg)
>>>>> {
>>>>> - if (!vreg)
>>>>> - return 0;
>>>>> - else if (vreg->unused)
>>>>> - return 0;
>>>>> - else
>>>>> - return ufshcd_config_vreg_load(hba->dev, vreg,
>>>>> - UFS_VREG_LPM_LOAD_UA);
>>>>> + return ufshcd_config_vreg_load(hba->dev, vreg, UFS_VREG_LPM_LOAD_UA);
>>>>> }
>>>>>
>>>>> static inline int ufshcd_config_vreg_hpm(struct ufs_hba *hba,
>>>>> struct ufs_vreg *vreg)
>>>>> {
>>>>> - if (!vreg)
>>>>> - return 0;
>>>>> - else if (vreg->unused)
>>>>> - return 0;
>>>>> - else
>>>>> - return ufshcd_config_vreg_load(hba->dev, vreg, vreg->max_uA);
>>>>> + return ufshcd_config_vreg_load(hba->dev, vreg, vreg->max_uA);
>>>>> }
>>>>>
>>>>> static int ufshcd_config_vreg(struct device *dev,
>>>>> @@ -7062,9 +7045,7 @@ static int ufshcd_enable_vreg(struct device *dev, struct ufs_vreg *vreg)
>>>>> {
>>>>> int ret = 0;
>>>>>
>>>>> - if (!vreg)
>>>>> - goto out;
>>>>> - else if (vreg->enabled || vreg->unused)
>>>>> + if (!vreg || vreg->enabled)
>>>>> goto out;
>>>>>
>>>>> ret = ufshcd_config_vreg(dev, vreg, true);
>>>>> @@ -7084,9 +7065,7 @@ static int ufshcd_disable_vreg(struct device *dev, struct ufs_vreg *vreg)
>>>>> {
>>>>> int ret = 0;
>>>>>
>>>>> - if (!vreg)
>>>>> - goto out;
>>>>> - else if (!vreg->enabled || vreg->unused)
>>>>> + if (!vreg || !vreg->enabled)
>>>>> goto out;
>>>>>
>>>>> ret = regulator_disable(vreg->reg);
>>>>> @@ -7192,36 +7171,6 @@ static int ufshcd_init_hba_vreg(struct ufs_hba *hba)
>>>>> return 0;
>>>>> }
>>>>>
>>>>> -static int ufshcd_set_vccq_rail_unused(struct ufs_hba *hba, bool unused)
>>>>> -{
>>>>> - int ret = 0;
>>>>> - struct ufs_vreg_info *info = &hba->vreg_info;
>>>>> -
>>>>> - if (!info)
>>>>> - goto out;
>>>>> - else if (!info->vccq)
>>>>> - goto out;
>>>>> -
>>>>> - if (unused) {
>>>>> - /* shut off the rail here */
>>>>> - ret = ufshcd_toggle_vreg(hba->dev, info->vccq, false);
>>>>> - /*
>>>>> - * Mark this rail as no longer used, so it doesn't get enabled
>>>>> - * later by mistake
>>>>> - */
>>>>> - if (!ret)
>>>>> - info->vccq->unused = true;
>>>>> - } else {
>>>>> - /*
>>>>> - * rail should have been already enabled hence just make sure
>>>>> - * that unused flag is cleared.
>>>>> - */
>>>>> - info->vccq->unused = false;
>>>>> - }
>>>>> -out:
>>>>> - return ret;
>>>>> -}
>>>>> -
>>>>> static int __ufshcd_setup_clocks(struct ufs_hba *hba, bool on,
>>>>> bool skip_ref_clk)
>>>>> {
>>>>>
>>>
>>>
>
>
Powered by blists - more mailing lists