Skip to contentSkip to conversation

52a518019c causes issue with Qualcomm UFSHC

DISCUSSION Asutosh Das 3 participants 4 messages Nov 21, 2022 – Nov 29, 2022 6 messages not in the archive SCSI

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): KL1PR03MB5836A982CAE0FE411743BF7283069@KL1PR03MB5836.apcprd03.prod.outlook.com

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): TYZPR03MB5825B4D5259FA22CCC876E7783089@TYZPR03MB5825.apcprd03.prod.outlook.com

#6 lore raw
Quoted 10 lines · Powen Kao (高伯文)

Hi Asutosh,

Reverting the patch doesn't sound feasible on MTK platform ☹

Quoted 1 line
Quoted 1 line

I don't think invoking a clock scaling notification during

Quoted 7 lines
Quoted 1 line
Quoted 1 line

ufshcd_host_reset_and_restore() sounds right to me.

But the point is that driver has the right to know that the clk is scaled no matter where ufshcd_scale_clks() is invoked, no?

Do you mind applying this patch on qcom driver to check on host status before further operation?

+ Mani, linux-scsi

Hello Powen
Thanks for the change. I will think of something to work-around the issue. However, I would like to point out that a change that breaks an existing driver must be fixed or reverted, not the other way around.

-asd

#7 lore raw

Hi Asutosh,

Sorry for the inconvenience, we will discuss internally to see if there's a better fix that meets both our needs.

-----Original Message-----
From: Asutosh Das <[email protected]>
Sent: Tuesday, November 22, 2022 6:57 AM
To: Powen Kao (高伯文) <[email protected]>
Cc: Bart Van Assche <[email protected]>; Stanley Chu (朱原陞) <[email protected]>; [email protected]; Peter Wang (王信友) <[email protected]>; Naomi Chu (朱詠田) <[email protected]>; Alice Chao (趙珮均) <[email protected]>; [email protected]; [email protected]; [email protected]
Subject: Re: 52a518019c causes issue with Qualcomm UFSHC

Quoted 10 lines · Powen Kao (高伯文)

Hi Asutosh,

Reverting the patch doesn't sound feasible on MTK platform ☹

Quoted 1 line
Quoted 1 line

I don't think invoking a clock scaling notification during

Quoted 7 lines
Quoted 1 line
Quoted 1 line

ufshcd_host_reset_and_restore() sounds right to me.

But the point is that driver has the right to know that the clk is scaled no matter where ufshcd_scale_clks() is invoked, no?

Do you mind applying this patch on qcom driver to check on host status before further operation?

+ Mani, linux-scsi

Hello Powen
Thanks for the change. I will think of something to work-around the issue.
However, I would like to point out that a change that breaks an existing driver must be fixed or reverted, not the other way around.

-asd

#8 lore raw
In reply to Asutosh Das
Quoted 17 lines · Asutosh Das
Quoted 16 lines · Powen Kao (高伯文)

Hi Asutosh,

Reverting the patch doesn't sound feasible on MTK platform ☹

Quoted 3 lines
Quoted 3 lines

I don't think invoking a clock scaling notification during

ufshcd_host_reset_and_restore() sounds right to me.

But the point is that driver has the right to know that the clk is scaled no matter where ufshcd_scale_clks() is invoked, no?

Quoted 9 lines
Quoted 1 line

Do you mind applying this patch on qcom driver to check on host status before further operation?

+ Mani, linux-scsi

Hello Powen
Thanks for the change. I will think of something to work-around the issue. However, I would like to point out that a change that breaks an existing driver must be fixed or reverted, not the other way around.

May I know what the actual issue is? I've tested v6.1-rc1 that has this change and didn't see any issues on SM8250, SDM845 and SM8450.

Thanks,
mani

Quoted 2 lines

-asd

Signature · 1 line
மணிவண்ணன் சதாசிவம்
#9 lore raw
Quoted 28 lines · Manivannan Sadhasivam
Quoted 17 lines · Asutosh Das
Quoted 16 lines · Powen Kao (高伯文)

Hi Asutosh,

Reverting the patch doesn't sound feasible on MTK platform ☹

Quoted 3 lines
Quoted 3 lines

I don't think invoking a clock scaling notification during

ufshcd_host_reset_and_restore() sounds right to me.

But the point is that driver has the right to know that the clk is scaled no matter where ufshcd_scale_clks() is invoked, no?

Quoted 9 lines
Quoted 1 line

Do you mind applying this patch on qcom driver to check on host status before further operation?

+ Mani, linux-scsi

Hello Powen
Thanks for the change. I will think of something to work-around the issue. However, I would like to point out that a change that breaks an existing driver must be fixed or reverted, not the other way around.

Quoted 6 lines

May I know what the actual issue is? I've tested v6.1-rc1 that has this change and didn't see any issues on SM8250, SDM845 and SM8450.

Thanks,
mani

Set rpm-lvl=5, enable scaling and let it enter runtime-suspend and then resume. You should see the CMD17 timeout.

-asd

Quoted 4 lines
Quoted 1 line
Signature · 1 line
மணிவண்ணன் சதாசிவம்