Skip to contentSkip to conversation

PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

PATCH Chris Clayton 3 participants 15 messages Aug 2, 2020 – Aug 22, 2020 Char and misc drivers

lore
Share & tools
lore
#0 lore raw

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card reader on my Intel NUC6CAYH box.

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp() was added to rtsx_pci_init_hw(). At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in rtsx_pci_init_ocp() the cardreader gets disabled.

I've avoided this by making excution code that results in the reader being disabled conditional on the device not being an RTS5229. Of course, other rtsxxx card readers may also be disabled by this bug. I don't have the knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Fixes bede03a579b3 ("misc: rtsx: Enable OCP for rts522a rts524a rts525a rts5260")

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card reader on my Intel NUC6CAYH box.

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp() was added to rtsx_pci_init_hw(). At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in rtsx_pci_init_ocp() the cardreader gets disabled.

I've avoided this by making excution code that results in the reader being disabled conditional on the device not being an RTS5229. Of course, other rtsxxx card readers may also be disabled by this bug. I don't have the knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Chris

drivers/misc/cardreader/rtsx_pcr.c renamed from drivers/misc/cardreader/rtsx_pcr.c.orig +7 -3
@@ -1200,9 +1200,13 @@ void rtsx_pci_init_ocp(struct rtsx_pcr *
 				SD_OCP_GLITCH_MASK, pcr->hw_param.ocp_glitch);
 			rtsx_pci_enable_ocp(pcr);
 		} else {
-			/* OC power down */
-			rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
-				OC_POWER_DOWN);
+			/* On (some?) Intel NUC platforms, this disables
+			 * the rts5229 cardreader, so don't do it
+			 */
+			if(!CHK_PCI_PID(pcr, 0x5229))
+				/* OC power down */
+				rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
+					OC_POWER_DOWN);
 		}
 	}
 }
#1 lore raw

Sorry, I should have said that the patch is against 5.7.12. It applies to upstream, but with offsets.

Quoted 31 lines · Chris Clayton

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card reader on my Intel NUC6CAYH box.

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp() was added to rtsx_pci_init_hw(). At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in rtsx_pci_init_ocp() the cardreader gets disabled.

I've avoided this by making excution code that results in the reader being disabled conditional on the device not being an RTS5229. Of course, other rtsxxx card readers may also be disabled by this bug. I don't have the knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Fixes bede03a579b3 ("misc: rtsx: Enable OCP for rts522a rts524a rts525a rts5260")

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card reader on my Intel NUC6CAYH box.

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp() was added to rtsx_pci_init_hw(). At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in rtsx_pci_init_ocp() the cardreader gets disabled.

I've avoided this by making excution code that results in the reader being disabled conditional on the device not being an RTS5229. Of course, other rtsxxx card readers may also be disabled by this bug. I don't have the knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Chris

#2 lore raw

Hi Chris,

We don’t think this is our bug...
This register(FPDCTL) write to OC_POWER_DOWN is for our power saving feature, not to disable the reader
In your case, we cannot reproduce this on our side that we mention before, we don’t have the platform(Intel NUC Tall Arches Canyon NUC6CAYH Celeron J345) to see this issue But we think this issue maybe only on this platform, our RTS5229 works well on the new kernel all platform that we have

Ricky

Quoted 56 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Monday, August 03, 2020 3:59 AM
To: LKML; 吳昊澄 Ricky; [email protected]; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Sorry, I should have said that the patch is against 5.7.12. It applies to upstream, but with offsets.

Quoted 1 line · Chris Clayton

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card

reader on my Intel NUC6CAYH box.

Quoted 2 lines

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp()

was added to rtsx_pci_init_hw().

Quoted 1 line

At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in

rtsx_pci_init_ocp() the cardreader

Quoted 3 lines

gets disabled.

I've avoided this by making excution code that results in the reader being

disabled conditional on the device

Quoted 1 line

not being an RTS5229. Of course, other rtsxxx card readers may also be

disabled by this bug. I don't have the

Quoted 5 lines

knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Fixes bede03a579b3 ("misc: rtsx: Enable OCP for rts522a rts524a rts525a

rts5260")

Quoted 5 lines

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card

reader on my Intel NUC6CAYH box.

Quoted 2 lines

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp()

was added to rtsx_pci_init_hw().

Quoted 1 line

At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in

rtsx_pci_init_ocp() the cardreader

Quoted 3 lines

gets disabled.

I've avoided this by making excution code that results in the reader being

disabled conditional on the device

Quoted 1 line

not being an RTS5229. Of course, other rtsxxx card readers may also be

disabled by this bug. I don't have the

Quoted 6 lines

knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Chris

------Please consider the environment before printing this e-mail.

#3 lore raw

Hi, Ricky

Quoted 8 lines · 吳昊澄 Ricky

Hi Chris,

We don’t think this is our bug...
This register(FPDCTL) write to OC_POWER_DOWN is for our power saving feature, not to disable the reader
In your case, we cannot reproduce this on our side that we mention before, we don’t have the platform(Intel NUC Tall Arches Canyon NUC6CAYH Celeron J345) to see this issue But we think this issue maybe only on this platform, our RTS5229 works well on the new kernel all platform that we have

Ricky

Perhaps I should have used the word regression rather than bug. I didn't buy the machine until earlier this year, but other people who have reported this problem have indicated that until bede03a579b3 was applied (during the 5.1 merge window), the driver supported the card reader on this on the Intel NUC boxes. I know that is true because I built and tested a 5.0 kernel and the card reader worked fine. I've also built and tested an 5.1-rc1 kernel and the card reader no longer works. Whether by design or by accident, the card reader worked and now it doesn't. That's a regression in my book!

Since you signed off the patch that caused the regression, I believe it is your bug.

Thanks.

Chris

Quoted 57 lines
Quoted 57 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Monday, August 03, 2020 3:59 AM
To: LKML; 吳昊澄 Ricky; [email protected]; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Sorry, I should have said that the patch is against 5.7.12. It applies to upstream, but with offsets.

Quoted 1 line · Chris Clayton

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card

reader on my Intel NUC6CAYH box.

Quoted 2 lines

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp()

was added to rtsx_pci_init_hw().

Quoted 1 line

At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in

rtsx_pci_init_ocp() the cardreader

Quoted 3 lines

gets disabled.

I've avoided this by making excution code that results in the reader being

disabled conditional on the device

Quoted 1 line

not being an RTS5229. Of course, other rtsxxx card readers may also be

disabled by this bug. I don't have the

Quoted 5 lines

knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Fixes bede03a579b3 ("misc: rtsx: Enable OCP for rts522a rts524a rts525a

rts5260")

Quoted 5 lines

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card

reader on my Intel NUC6CAYH box.

Quoted 2 lines

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp()

was added to rtsx_pci_init_hw().

Quoted 1 line

At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in

rtsx_pci_init_ocp() the cardreader

Quoted 3 lines

gets disabled.

I've avoided this by making excution code that results in the reader being

disabled conditional on the device

Quoted 1 line

not being an RTS5229. Of course, other rtsxxx card readers may also be

disabled by this bug. I don't have the

Quoted 6 lines

knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Chris

------Please consider the environment before printing this e-mail.

#4 lore raw
In reply to Chris Clayton

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN, OC_POWER_DOWN);
This register operation saved power under 1mA, so if do not care the 1mA power we can have a patch to remove it, make compatible with NUC6 We tested others our card reader that remove it, we did not see any side effect

Hi Greg k-h,

Do you have any comments?

thanks

Ricky

Quoted 56 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Monday, August 03, 2020 3:59 AM
To: LKML; 吳昊澄 Ricky; [email protected]; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Sorry, I should have said that the patch is against 5.7.12. It applies to upstream, but with offsets.

Quoted 1 line · Chris Clayton

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card

reader on my Intel NUC6CAYH box.

Quoted 2 lines

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp()

was added to rtsx_pci_init_hw().

Quoted 1 line

At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in

rtsx_pci_init_ocp() the cardreader

Quoted 3 lines

gets disabled.

I've avoided this by making excution code that results in the reader being

disabled conditional on the device

Quoted 1 line

not being an RTS5229. Of course, other rtsxxx card readers may also be

disabled by this bug. I don't have the

Quoted 5 lines

knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Fixes bede03a579b3 ("misc: rtsx: Enable OCP for rts522a rts524a rts525a

rts5260")

Quoted 5 lines

bede03a579b3 introduced a bug which leaves the rts5229 PCI Express card

reader on my Intel NUC6CAYH box.

Quoted 2 lines

The bug is in drivers/misc/cardreader/rtsx_pcr.c. A call to rtsx_pci_init_ocp()

was added to rtsx_pci_init_hw().

Quoted 1 line

At the call point, pcr->ops->init_ocp is NULL and pcr->option.ocp_en is 0, so in

rtsx_pci_init_ocp() the cardreader

Quoted 3 lines

gets disabled.

I've avoided this by making excution code that results in the reader being

disabled conditional on the device

Quoted 1 line

not being an RTS5229. Of course, other rtsxxx card readers may also be

disabled by this bug. I don't have the

Quoted 6 lines

knowledge to address that, so I'll leave to the driver maintainers.

The patch to avoid the bug is attached.

Chris

------Please consider the environment before printing this e-mail.

#5 lore raw
Quoted 9 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN, OC_POWER_DOWN);
This register operation saved power under 1mA, so if do not care the 1mA power we can have a patch to remove it, make compatible with NUC6 We tested others our card reader that remove it, we did not see any side effect

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

greg k-h

#6 lore raw
Quoted 14 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 4 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN, OC_POWER_DOWN); This register operation saved power under 1mA, so if do not care the 1mA

Quoted 9 lines

power we can have a patch to remove it, make compatible with NUC6

Quoted 5 lines

We tested others our card reader that remove it, we did not see any side effect

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but it will have more compatibility

Quoted 3 lines

greg k-h

------Please consider the environment before printing this e-mail.

#7 lore raw
Quoted 17 lines · 吳昊澄 Ricky
Quoted 17 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 4 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN, OC_POWER_DOWN); This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 2 lines

We tested others our card reader that remove it, we did not see any side effect

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but it will have more compatibility

I do not understand at all, sorry.

greg k-h

#8 lore raw
In reply to 吳昊澄 Ricky
Quoted 17 lines · 吳昊澄 Ricky
Quoted 17 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 4 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN, OC_POWER_DOWN); This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 2 lines

We tested others our card reader that remove it, we did not see any side effect

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but it will have more compatibility

Ricky,

I don't understand why you want to completely remove the code that sets up the 1mA power saving. That code was there even before your patch (bede03a579b3b4a036003c4862cc1baa4ddc351f), so I assume it benefits some of the Realtek card readers. Before your patch however, rtsx_pci_init_ocp() was not called for the rts5229 reader, but the patch introduced an unconditional call to that function into rtsx_pci_init_hw(), which is run for the rts5229. That is what now disables the card reader.

Now, I don't know whether other cards are affected, although I don't recall seeing any reported as I searched the kernel and ubuntu bugzillas for any analysis of the problem. I know this is not what the patch I sent does, but having thought about it more, seems to me that the simplest fix is to skip the new call to rtsx_pci_init_ocp() if the reader is an rts5229.

If you agree, I can prepare a patch and send it to you. Whatever the solution is, it will also be needed in the stable kernels later than 5.0.

Chris

Quoted 3 lines
Quoted 3 lines

greg k-h

------Please consider the environment before printing this e-mail.

#9 lore raw
Quoted 53 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 4:51 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 5 lines · 吳昊澄 Ricky
Quoted 5 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected];

Arnd

Quoted 8 lines
Quoted 8 lines

Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 3 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN,

OC_POWER_DOWN);

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 1 line

We tested others our card reader that remove it, we did not see any side

effect

Quoted 8 lines
Quoted 7 lines
Quoted 4 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but it

will have more compatibility

Quoted 1 line

Ricky,

I don't understand why you want to completely remove the code that sets up the 1mA power saving. That code was there
even before your patch (bede03a579b3b4a036003c4862cc1baa4ddc351f), so I assume it benefits some of the Realtek card
readers. Before your patch however, rtsx_pci_init_ocp() was not called for the rts5229 reader, but the patch introduced
an unconditional call to that function into rtsx_pci_init_hw(), which is run for the rts5229. That is what now disables
the card reader.

Quoted 7 lines

Now, I don't know whether other cards are affected, although I don't recall seeing any reported as I searched the kernel
and ubuntu bugzillas for any analysis of the problem. I know this is not what the patch I sent does, but having thought
about it more, seems to me that the simplest fix is to skip the new call to rtsx_pci_init_ocp() if the reader is an rts5229.

Because we are thinking about if others our card reader that not belong A series(my ocp patch coverage) also on NUC6 platform maybe have the same problem...

Quoted 4 lines

If you agree, I can prepare a patch and send it to you. Whatever the solution is, it will also be needed in the stable
kernels later than 5.0.

OK, I agree your opinion, for now can only patch rts5229 first make NUC6 user can work well

Thank you

Quoted 4 lines

Chris

Quoted 3 lines
Quoted 3 lines

greg k-h

------Please consider the environment before printing this e-mail.

#10 lore raw
Quoted 61 lines · 吳昊澄 Ricky
Quoted 53 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 4:51 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 5 lines · 吳昊澄 Ricky
Quoted 5 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected];

Arnd

Quoted 8 lines
Quoted 8 lines

Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 3 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN,

OC_POWER_DOWN);

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 1 line

We tested others our card reader that remove it, we did not see any side

effect

Quoted 8 lines
Quoted 7 lines
Quoted 4 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but it

will have more compatibility

Quoted 1 line

Ricky,

I don't understand why you want to completely remove the code that sets up the 1mA power saving. That code was there
even before your patch (bede03a579b3b4a036003c4862cc1baa4ddc351f), so I assume it benefits some of the Realtek card
readers. Before your patch however, rtsx_pci_init_ocp() was not called for the rts5229 reader, but the patch introduced
an unconditional call to that function into rtsx_pci_init_hw(), which is run for the rts5229. That is what now disables
the card reader.

Quoted 7 lines

Now, I don't know whether other cards are affected, although I don't recall seeing any reported as I searched the kernel
and ubuntu bugzillas for any analysis of the problem. I know this is not what the patch I sent does, but having thought
about it more, seems to me that the simplest fix is to skip the new call to rtsx_pci_init_ocp() if the reader is an rts5229.

Quoted 2 lines

Because we are thinking about if others our card reader that not belong A series(my ocp patch coverage) also on NUC6 platform maybe have the same problem...

OK. What if we do make the new call but only for the card readers that are in the A series? Are they the ones that have PID_5nnn defines in include/linux/rtsx_pci.h? Or is there another simple way of identifying that a reader is a member of the A series?

I'm thinking of something like:
static bool rtsx_pci_is_series_A(pcr)

{
	unsigned short device = pcr->pci->device;

	return device == PID524A || device == PID_5249 || device == PID_5250 || device == PID_525A
			|| device == PID_525A || device == PID_5260 || device == PID_5261;
}

then in rtsx_pci_init_hw() change the unconditional call to:

	if rtsx_pci_is_series_A(pcr)
		rtsx_pci_init_ocp();

Does that seem OK?

Quoted 13 lines
Quoted 4 lines

If you agree, I can prepare a patch and send it to you. Whatever the solution is, it will also be needed in the stable
kernels later than 5.0.

OK, I agree your opinion, for now can only patch rts5229 first make NUC6 user can work well

Thank you

Quoted 4 lines

Chris

Quoted 3 lines
Quoted 3 lines

greg k-h

------Please consider the environment before printing this e-mail.

#11 lore raw
Quoted 101 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 7:52 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 11 lines · 吳昊澄 Ricky
Quoted 11 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 4:51 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

On 04/08/2020 09:08, 吳昊澄 Ricky wrote:

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected];

Arnd

Quoted 2 lines
Quoted 2 lines

Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader

on

Quoted 12 lines
Quoted 12 lines
Quoted 6 lines
Quoted 6 lines

Intel NUC boxes

Quoted 3 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN,

OC_POWER_DOWN);

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 1 line

We tested others our card reader that remove it, we did not see any side

effect

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but

it

Quoted 6 lines
Quoted 6 lines

will have more compatibility

Quoted 1 line

Ricky,

I don't understand why you want to completely remove the code that sets up

the

Quoted 6 lines
Quoted 6 lines

1mA power saving. That code was there
even before your patch (bede03a579b3b4a036003c4862cc1baa4ddc351f), so I assume it benefits some of the Realtek card
readers. Before your patch however, rtsx_pci_init_ocp() was not called for the rts5229 reader, but the patch introduced
an unconditional call to that function into rtsx_pci_init_hw(), which is run for

the

Quoted 6 lines
Quoted 6 lines

rts5229. That is what now disables
the card reader.

Now, I don't know whether other cards are affected, although I don't recall seeing any reported as I searched the kernel
and ubuntu bugzillas for any analysis of the problem. I know this is not what

the

Quoted 6 lines
Quoted 4 lines

patch I sent does, but having thought
about it more, seems to me that the simplest fix is to skip the new call to rtsx_pci_init_ocp() if the reader is an rts5229.

Because we are thinking about if others our card reader that not belong A

series(my ocp patch coverage) also on NUC6 platform maybe have the same problem...

Quoted 1 line

OK. What if we do make the new call but only for the card readers that are in the A series? Are they the ones that have
PID_5nnn defines in include/linux/rtsx_pci.h? Or is there another simple way of identifying that a reader is a member of
the A series?

I'm thinking of something like:
static bool rtsx_pci_is_series_A(pcr)

{
	unsigned short device = pcr->pci->device;

	return device == PID524A || device == PID_5249 || device == PID_5250 ||
device == PID_525A
			|| device == PID_525A || device == PID_5260 || device ==

PID_5261;
}

Quoted 7 lines

then in rtsx_pci_init_hw() change the unconditional call to:

	if rtsx_pci_is_series_A(pcr)
		rtsx_pci_init_ocp();

Does that seem OK?

Previously, I want to remove

else {
		/* OC power down */
		rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
			OC_POWER_DOWN);
}

Because in our A-series card Reader we already assigned option->ocp_en to 1 in self init_params() , this is an easy way to fix this problem

Quoted 15 lines
Quoted 1 line
Quoted 1 line

If you agree, I can prepare a patch and send it to you. Whatever the solution is,

it

Quoted 5 lines
Quoted 3 lines

will also be needed in the stable
kernels later than 5.0.

OK, I agree your opinion, for now can only patch rts5229 first make NUC6 user

can work well

Quoted 7 lines

Thank you

Quoted 4 lines

Chris

Quoted 3 lines
Quoted 3 lines

greg k-h

------Please consider the environment before printing this e-mail.

#12 lore raw

Thanks, Ricky.

Quoted 108 lines · 吳昊澄 Ricky
Quoted 101 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 7:52 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 11 lines · 吳昊澄 Ricky
Quoted 11 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 4:51 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

On 04/08/2020 09:08, 吳昊澄 Ricky wrote:

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected];

Arnd

Quoted 2 lines
Quoted 2 lines

Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader

on

Quoted 12 lines
Quoted 12 lines
Quoted 6 lines
Quoted 6 lines

Intel NUC boxes

Quoted 3 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN,

OC_POWER_DOWN);

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 1 line

We tested others our card reader that remove it, we did not see any side

effect

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but

it

Quoted 6 lines
Quoted 6 lines

will have more compatibility

Quoted 1 line

Ricky,

I don't understand why you want to completely remove the code that sets up

the

Quoted 6 lines
Quoted 6 lines

1mA power saving. That code was there
even before your patch (bede03a579b3b4a036003c4862cc1baa4ddc351f), so I assume it benefits some of the Realtek card
readers. Before your patch however, rtsx_pci_init_ocp() was not called for the rts5229 reader, but the patch introduced
an unconditional call to that function into rtsx_pci_init_hw(), which is run for

the

Quoted 6 lines
Quoted 6 lines

rts5229. That is what now disables
the card reader.

Now, I don't know whether other cards are affected, although I don't recall seeing any reported as I searched the kernel
and ubuntu bugzillas for any analysis of the problem. I know this is not what

the

Quoted 6 lines
Quoted 4 lines

patch I sent does, but having thought
about it more, seems to me that the simplest fix is to skip the new call to rtsx_pci_init_ocp() if the reader is an rts5229.

Because we are thinking about if others our card reader that not belong A

series(my ocp patch coverage) also on NUC6 platform maybe have the same problem...

Quoted 1 line

OK. What if we do make the new call but only for the card readers that are in the A series? Are they the ones that have
PID_5nnn defines in include/linux/rtsx_pci.h? Or is there another simple way of identifying that a reader is a member of
the A series?

I'm thinking of something like:
static bool rtsx_pci_is_series_A(pcr)

{
	unsigned short device = pcr->pci->device;

	return device == PID524A || device == PID_5249 || device == PID_5250 ||
device == PID_525A
			|| device == PID_525A || device == PID_5260 || device ==

PID_5261;
}

Quoted 7 lines

then in rtsx_pci_init_hw() change the unconditional call to:

	if rtsx_pci_is_series_A(pcr)
		rtsx_pci_init_ocp();

Does that seem OK?

Quoted 8 lines

Previously, I want to remove

else {
		/* OC power down */
		rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
			OC_POWER_DOWN);
}

Because in our A-series card Reader we already assigned option->ocp_en to 1 in self init_params() , this is an easy way to fix this problem

Ah, OK. I'll prepare the patch and send it to you once I've tested it.

Chris

#13 lore raw

Hi Ricky

Quoted 120 lines · Chris Clayton

Thanks, Ricky.

Quoted 108 lines · 吳昊澄 Ricky
Quoted 101 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 7:52 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

Quoted 11 lines · 吳昊澄 Ricky
Quoted 11 lines

-----Original Message-----
From: Chris Clayton [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 4:51 PM
To: 吳昊澄 Ricky; [email protected]
Cc: LKML; [email protected]; [email protected]; Arnd Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader on Intel NUC boxes

On 04/08/2020 09:08, 吳昊澄 Ricky wrote:

Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, August 04, 2020 3:49 PM
To: 吳昊澄 Ricky
Cc: Chris Clayton; LKML; [email protected]; [email protected];

Arnd

Quoted 2 lines
Quoted 2 lines

Bergmann
Subject: Re: PATCH: rtsx_pci driver - don't disable the rts5229 card reader

on

Quoted 12 lines
Quoted 12 lines
Quoted 6 lines
Quoted 6 lines

Intel NUC boxes

Quoted 3 lines · 吳昊澄 Ricky

Hi Chris,

rtsx_pci_write_register(pcr, FPDTL, OC_POWER_DOWN,

OC_POWER_DOWN);

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

This register operation saved power under 1mA, so if do not care the 1mA

power we can have a patch to remove it, make compatible with NUC6

Quoted 1 line

We tested others our card reader that remove it, we did not see any side

effect

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines

Hi Greg k-h,

Do you have any comments?

comments on what? I don't know what you are responding to here, sorry.

Can we have a patch to kernel for NUC6? It may cause more power(1mA) but

it

Quoted 6 lines
Quoted 6 lines

will have more compatibility

Quoted 1 line

Ricky,

I don't understand why you want to completely remove the code that sets up

the

Quoted 6 lines
Quoted 6 lines

1mA power saving. That code was there
even before your patch (bede03a579b3b4a036003c4862cc1baa4ddc351f), so I assume it benefits some of the Realtek card
readers. Before your patch however, rtsx_pci_init_ocp() was not called for the rts5229 reader, but the patch introduced
an unconditional call to that function into rtsx_pci_init_hw(), which is run for

the

Quoted 6 lines
Quoted 6 lines

rts5229. That is what now disables
the card reader.

Now, I don't know whether other cards are affected, although I don't recall seeing any reported as I searched the kernel
and ubuntu bugzillas for any analysis of the problem. I know this is not what

the

Quoted 6 lines
Quoted 4 lines

patch I sent does, but having thought
about it more, seems to me that the simplest fix is to skip the new call to rtsx_pci_init_ocp() if the reader is an rts5229.

Because we are thinking about if others our card reader that not belong A

series(my ocp patch coverage) also on NUC6 platform maybe have the same problem...

Quoted 1 line

OK. What if we do make the new call but only for the card readers that are in the A series? Are they the ones that have
PID_5nnn defines in include/linux/rtsx_pci.h? Or is there another simple way of identifying that a reader is a member of
the A series?

I'm thinking of something like:
static bool rtsx_pci_is_series_A(pcr)

{
	unsigned short device = pcr->pci->device;

	return device == PID524A || device == PID_5249 || device == PID_5250 ||
device == PID_525A
			|| device == PID_525A || device == PID_5260 || device ==

PID_5261;
}

Quoted 7 lines

then in rtsx_pci_init_hw() change the unconditional call to:

	if rtsx_pci_is_series_A(pcr)
		rtsx_pci_init_ocp();

Does that seem OK?

Quoted 8 lines

Previously, I want to remove

else {
		/* OC power down */
		rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
			OC_POWER_DOWN);
}

Because in our A-series card Reader we already assigned option->ocp_en to 1 in self init_params() , this is an easy way to fix this problem

Quoted 2 lines

Ah, OK. I'll prepare the patch and send it to you once I've tested it.

Please see the patch included below. As you suggested, it removes the code that does the OC power down on card readers that are not members of your A series. The patch is against a fresh pull of Linus's tree this morning (v5.8-2768-g4da9f3302615).

I've tested the resultant kernel on my Intel NUC6CAYH box (which contains an NUC66AYB board) and the rts5229 works fine. I've also tested it on my laptop which also has a card reader supported by the rtsx_pci driver (an RTL8411B) and that also works fine.

As I mentioned yesterday, I think it's a candidate for the 5.4 ans 5.7 stable series.

Thanks for your patience!

Chris

drivers/misc/cardreader/rtsx_pcr.c 2020-08-05 07:11:05.568074278 +0100 renamed from drivers/misc/cardreader/rtsx_pcr.c 2020-08-05 07:10:21.752072515 +0100 +0 -4
@@ -1172,10 +1172,6 @@ void rtsx_pci_init_ocp(struct rtsx_pcr *
                        rtsx_pci_write_register(pcr, REG_OCPGLITCH,
                                SD_OCP_GLITCH_MASK, pcr->hw_param.ocp_glitch);
                        rtsx_pci_enable_ocp(pcr);
-               } else {
-                       /* OC power down */
-                       rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
-                               OC_POWER_DOWN);
                }
        }
 }
#14 lore raw

Hi Ricky.

Quoted 1 line · Chris Clayton

Hi Ricky

<snip>

Quoted 25 lines
Quoted 3 lines

Ah, OK. I'll prepare the patch and send it to you once I've tested it.

Please see the patch included below. As you suggested, it removes the code that does the OC power down on card readers that are not members of your A series. The patch is against a fresh pull of Linus's tree this morning (v5.8-2768-g4da9f3302615).

I've tested the resultant kernel on my Intel NUC6CAYH box (which contains an NUC66AYB board) and the rts5229 works fine. I've also tested it on my laptop which also has a card reader supported by the rtsx_pci driver (an RTL8411B) and that also works fine.

As I mentioned yesterday, I think it's a candidate for the 5.4 ans 5.7 stable series.

Thanks for your patience!

Chris

drivers/misc/cardreader/rtsx_pcr.c 2020-08-05 07:11:05.568074278 +0100 void rtsx_pci_init_ocp(struct rtsx_pcr * · line 1177

Quoted 10 lines
drivers/misc/cardreader/rtsx_pcr.c 2020-08-05 07:11:05.568074278 +0100 renamed from drivers/misc/cardreader/rtsx_pcr.c 2020-08-05 07:10:21.752072515 +0100 +0 -4
@@ -1172,10 +1172,6 @@ void rtsx_pci_init_ocp(struct rtsx_pcr *
                        rtsx_pci_write_register(pcr, REG_OCPGLITCH,
                                SD_OCP_GLITCH_MASK, pcr->hw_param.ocp_glitch);
                        rtsx_pci_enable_ocp(pcr);
-               } else {
-                       /* OC power down */
-                       rtsx_pci_write_register(pcr, FPDCTL, OC_POWER_DOWN,
-                               OC_POWER_DOWN);
                }
        }
 }

Is there some problem with my patch? If you are too busy to deal with it, perhaps I can submit directly it to Greg/Arnd.

Thanks

Chris