usb: dwc3: gadget: Enable suspend events
Subsystems (2)
PATCH 1/2 usb: dwc3: gadget: Enable suspend events
commit 72704f876f50 ("dwc3: gadget: Implement the suspend entry event handler") introduced (nearly 5 years ago!) an interrupt handler for U3/L1-L2 suspend events. The problem is that these events aren't currently enabled in the DEVTEN register so the handler is never even invoked. Fix this simply by enabling the corresponding bit in dwc3_gadget_enable_irq() using the same revision check as found in the handler.
--- drivers/usb/dwc3/gadget.c | 4 ++++ 1 file changed, 4 insertions(+)
drivers/usb/dwc3/gadget.c +4 -0
@@ -2323,6 +2323,10 @@ static void dwc3_gadget_enable_irq(struct dwc3 *dwc) if (DWC3_VER_IS_PRIOR(DWC3, 250A)) reg |=DWC3_DEVTEN_ULSTCNGEN ; +/* On 2.30a and above this bit enables U3/L2-L1 Suspend Events */ + if (!DWC3_VER_IS_PRIOR(DWC3, 230A)) + reg |=DWC3_DEVTEN_EOPFEN ; +dwc3_writel (dwc->regs,DWC3_DEVTEN , reg); }
Signature · 1 line
2.24.0
PATCH 2/2 usb: dwc3: gadget: Rename EOPF event macros to Suspend
The device event corresponding to End of Periodic Frame is only found on older IP revisions (2.10a and prior, according to a cursory SNPS databook search). On revisions 2.30a and newer, including DWC3.1, the same event value and corresponding DEVTEN bit were repurposed to indicate that the link has gone into suspend state (U3 or L2/L1).
EOPF events had never been enabled before in this driver, and going forward we expect current and future DWC3-based devices won't likely to be using such old DWC3 IP revisions either. Hence rather than keeping the deprecated EOPF macro names let's rename them to indicate their usage for suspend events.
--- drivers/usb/dwc3/core.h | 6 +++--- drivers/usb/dwc3/debug.h | 8 ++++---- drivers/usb/dwc3/gadget.c | 4 ++-- 3 files changed, 9 insertions(+), 9 deletions(-)
drivers/usb/dwc3/core.h +3 -3
@@ -57,7 +57,7 @@#define DWC3_DEVICE_EVENT_LINK_STATUS_CHANGE 3#define DWC3_DEVICE_EVENT_WAKEUP 4#define DWC3_DEVICE_EVENT_HIBER_REQ 5 -#define DWC3_DEVICE_EVENT_EOPF 6 +#define DWC3_DEVICE_EVENT_SUSPEND 6#define DWC3_DEVICE_EVENT_SOF 7#define DWC3_DEVICE_EVENT_ERRATIC_ERROR 9#define DWC3_DEVICE_EVENT_CMD_CMPL 10
@@ -460,7 +460,7 @@#define DWC3_DEVTEN_CMDCMPLTEN BIT(10)#define DWC3_DEVTEN_ERRTICERREN BIT(9)#define DWC3_DEVTEN_SOFEN BIT(7) -#define DWC3_DEVTEN_EOPFEN BIT(6) +#define DWC3_DEVTEN_U3L2L1SUSPEN BIT(6)#define DWC3_DEVTEN_HIBERNATIONREQEVTEN BIT(5)#define DWC3_DEVTEN_WKUPEVTEN BIT(4)#define DWC3_DEVTEN_ULSTCNGEN BIT(3)
@@ -1374,7 +1374,7 @@ struct dwc3_event_depevt {* 3 - ULStChng * 4 - WkUpEvt * 5 - Reserved -* 6 - EOPF +* 6 - Suspend (EOPF on revisions 2.10a and prior) * 7 - SOF * 8 - Reserved * 9 - ErrticErr
drivers/usb/dwc3/debug.h +4 -4
@@ -221,8 +221,8 @@ static inline const char *dwc3_gadget_event_string(char *str, size_t size,snprintf (str, size,"WakeUp [%s]" ,dwc3_gadget_link_string (state));break ; -case DWC3_DEVICE_EVENT_EOPF : -snprintf (str, size,"End-Of-Frame [%s]" , +case DWC3_DEVICE_EVENT_SUSPEND : +snprintf (str, size,"Suspend [%s]" ,dwc3_gadget_link_string (state));break ;case DWC3_DEVICE_EVENT_SOF :
@@ -353,8 +353,8 @@ static inline const char *dwc3_gadget_event_type_string(u8 event)return "Wake-Up" ;case DWC3_DEVICE_EVENT_HIBER_REQ :return "Hibernation" ; -case DWC3_DEVICE_EVENT_EOPF : -return "End of Periodic Frame" ; +case DWC3_DEVICE_EVENT_SUSPEND : +return "Suspend" ;case DWC3_DEVICE_EVENT_SOF :return "Start of Frame" ;case DWC3_DEVICE_EVENT_ERRATIC_ERROR :
drivers/usb/dwc3/gadget.c +2 -2
@@ -2325,7 +2325,7 @@ static void dwc3_gadget_enable_irq(struct dwc3 *dwc)/* On 2.30a and above this bit enables U3/L2-L1 Suspend Events */ if (!DWC3_VER_IS_PRIOR(DWC3, 230A)) - reg |=DWC3_DEVTEN_EOPFEN ; + reg |=DWC3_DEVTEN_U3L2L1SUSPEN ;dwc3_writel (dwc->regs,DWC3_DEVTEN , reg); }
@@ -3744,7 +3744,7 @@ static void dwc3_gadget_interrupt(struct dwc3 *dwc,case DWC3_DEVICE_EVENT_LINK_STATUS_CHANGE :dwc3_gadget_linksts_change_interrupt (dwc, event->event_info);break ; -case DWC3_DEVICE_EVENT_EOPF : +case DWC3_DEVICE_EVENT_SUSPEND :/* It changed to be suspend event for version 2.30a and above */ if (!DWC3_VER_IS_PRIOR(DWC3, 230A)) {/*
Signature · 1 line
2.24.0
Re: [PATCH 2/2] usb: dwc3: gadget: Rename EOPF event macros to Suspend
Hi,
Quoted 2 lines · Jack Pham
The device event corresponding to End of Periodic Frame is only found on older IP revisions (2.10a and prior, according to a
you're reading databook for dwc3.1, right? Remember the original support for the driver was on dwc3. This was always called EOPF back then. We should maintain the name.
Quoted 8 lines
cursory SNPS databook search). On revisions 2.30a and newer, including DWC3.1, the same event value and corresponding DEVTEN bit were repurposed to indicate that the link has gone into suspend state (U3 or L2/L1).
EOPF events had never been enabled before in this driver, and going forward we expect current and future DWC3-based devices won't likely to be using such old DWC3 IP revisions either.
We still have original omap5 devices, running on revision 1.73a around. They'll remain supported for the time being.
Quoted 2 lines
Hence rather than keeping the deprecated EOPF macro names let's rename them to indicate their usage for suspend events.
what do we gain from this change? I mean, in practice, what changes? nothing realy, so why should we apply this?
Signature · 1 line
balbi
Re: [PATCH 2/2] usb: dwc3: gadget: Rename EOPF event macros to Suspend
Hi Felipe,
Quoted 7 lines · Felipe Balbi
Quoted 2 lines · Jack Pham
The device event corresponding to End of Periodic Frame is only found on older IP revisions (2.10a and prior, according to a
you're reading databook for dwc3.1, right? Remember the original support for the driver was on dwc3. This was always called EOPF back then. We should maintain the name.
I've looked through several revisions of the databook for both dwc3 and dwc3.1. From what I can tell EOPF was nixed starting in DWC3 (not 3.1) revision 2.20a. DWC3 revision 2.30a re-introduced event #6 for USB suspend. And judging from the IP revision list in core.h, DWC3 is now up to 3.30a (DWC3_REVISION_330A), so from number alone there are about as many revisions that have this bit as EOPF as there are that use it for SUSPEND. This carries over to DWC3.1 as well (not sure about DWC3.2) so in fact there are probably more revisions of IP that no longer use EOPF.
Hi Thinh, I'm wondering if you could please help corroborate the history of this bit, and confirm whether it is also used as Suspend entry in DWC 3.2 IPs?
But I don't want to make it seem that I'm using revision history as a gauge of how many real devices out there support EOPF vs Suspend. That figure we'll never truly know.
Quoted 9 lines
Quoted 8 lines
cursory SNPS databook search). On revisions 2.30a and newer, including DWC3.1, the same event value and corresponding DEVTEN bit were repurposed to indicate that the link has gone into suspend state (U3 or L2/L1).
EOPF events had never been enabled before in this driver, and going forward we expect current and future DWC3-based devices won't likely to be using such old DWC3 IP revisions either.
Quoted 8 lines
We still have original omap5 devices, running on revision 1.73a around. They'll remain supported for the time being.
Quoted 2 lines
Hence rather than keeping the deprecated EOPF macro names let's rename them to indicate their usage for suspend events.
what do we gain from this change? I mean, in practice, what changes? nothing realy, so why should we apply this?
I'm saying since this macro has never really been used to enable any kind of event handling specifically for "End Of Periodic Frame", that there is not much utility in keeping the name as EOPF. Instead as I explained in patch 1, the same bit/event is used on newer revisions for USB Suspend entry so assuming you accept that, then the purpose of this follow-on patch is simply to make the code more readable by renaming the macro to fit its usage.
Thanks,
Jack
Signature · 2 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project
Thinh Nguyen
maintainerRe: [PATCH 2/2] usb: dwc3: gadget: Rename EOPF event macros to Suspend
Quoted 21 lines · Jack Pham
Hi Felipe,
Quoted 7 lines · Felipe Balbi
Quoted 2 lines · Jack Pham
The device event corresponding to End of Periodic Frame is only found on older IP revisions (2.10a and prior, according to a
you're reading databook for dwc3.1, right? Remember the original support for the driver was on dwc3. This was always called EOPF back then. We should maintain the name.
I've looked through several revisions of the databook for both dwc3 and dwc3.1. From what I can tell EOPF was nixed starting in DWC3 (not 3.1) revision 2.20a. DWC3 revision 2.30a re-introduced event #6 for USB suspend. And judging from the IP revision list in
core.h, DWC3 is now up to 3.30a (DWC3_REVISION_330A), so from number alone there are about as many revisions that have this bit as EOPF as there are that use it for SUSPEND. This carries over to DWC3.1 as well (not sure about DWC3.2) so in fact there are probably more revisions of IP that no longer use EOPF.
Quoted 3 lines
Hi Thinh, I'm wondering if you could please help corroborate the history of this bit, and confirm whether it is also used as Suspend entry in DWC 3.2 IPs?
DWC_usb32 IP is also the same as DWC_usb31 and uses it for suspend event.
BR,
Thinh
Quoted 34 lines
But I don't want to make it seem that I'm using revision history as a gauge of how many real devices out there support EOPF vs Suspend. That figure we'll never truly know.
Quoted 9 lines
Quoted 8 lines
cursory SNPS databook search). On revisions 2.30a and newer, including DWC3.1, the same event value and corresponding DEVTEN bit were repurposed to indicate that the link has gone into suspend state (U3 or L2/L1).
EOPF events had never been enabled before in this driver, and going forward we expect current and future DWC3-based devices won't likely to be using such old DWC3 IP revisions either.
Quoted 8 lines
We still have original omap5 devices, running on revision 1.73a around. They'll remain supported for the time being.
Quoted 2 lines
Hence rather than keeping the deprecated EOPF macro names let's rename them to indicate their usage for suspend events.
what do we gain from this change? I mean, in practice, what changes? nothing realy, so why should we apply this?
I'm saying since this macro has never really been used to enable any kind of event handling specifically for "End Of Periodic Frame", that there is not much utility in keeping the name as EOPF. Instead as I explained in patch 1, the same bit/event is used on newer revisions for USB Suspend entry so assuming you accept that, then the purpose of this follow-on patch is simply to make the code more readable by renaming the macro to fit its usage.
Thanks,
Jack
Re: [PATCH 2/2] usb: dwc3: gadget: Rename EOPF event macros to Suspend
In reply to Jack PhamQuoted 7 lines · Jack Pham
The device event corresponding to End of Periodic Frame is only found on older IP revisions (2.10a and prior, according to a cursory SNPS databook search). On revisions 2.30a and newer, including DWC3.1, the same event value and corresponding DEVTEN bit were repurposed to indicate that the link has gone into suspend state (U3 or L2/L1).
Quoted 7 lines
EOPF events had never been enabled before in this driver, and going forward we expect current and future DWC3-based devices won't likely to be using such old DWC3 IP revisions either. Hence rather than keeping the deprecated EOPF macro names let's rename them to indicate their usage for suspend events.
Signed-off-by Jack Pham <[email protected]>
Signature · 1 line
balbi
Hi,
Quoted 3 lines · Jack Pham
commit
72704f876f50("dwc3: gadget: Implement the suspend entry event handler") introduced (nearly 5 years ago!) an interrupt handler for U3/L1-L2 suspend events. The problem is that these events aren't
look deeper. They *were* enabled. We've removed because, as it turns out, they just add a TON of interrupts and don't give us much extra information. The only reason why the state change interrupts are still there is because of a known silicon bug in versions < 2.50a. That's all documented in the driver itself.
For anything that works, we *don't* want link state change interrupts.
Quoted 4 lines
currently enabled in the DEVTEN register so the handler is never even invoked. Fix this simply by enabling the corresponding bit in
dwc3_gadget_enable_irq()using the same revision check as found in the handler.
More importantly, *why* do you think you need these interrupts?
Quoted 11 lines
Fixes 72704f876f50 ("dwc3: gadget: Implement the suspend entry event handler")Signed-off-by Jack Pham <[email protected]>--- drivers/usb/dwc3/gadget.c | 4 ++++ 1 file changed, 4 insertions(+)drivers/usb/dwc3/gadget.c +0
-0
drivers/usb/dwc3/gadget.c static void dwc3_gadget_enable_irq(struct dwc3 *dwc) · line 2328
Quoted 6 lines
drivers/usb/dwc3/gadget.c +3
-0@@ -2323,6 +2323,10 @@ static void dwc3_gadget_enable_irq(struct dwc3 *dwc) if (DWC3_VER_IS_PRIOR(DWC3, 250A)) reg |=DWC3_DEVTEN_ULSTCNGEN ; +/* On 2.30a and above this bit enables U3/L2-L1 Suspend Events */ + if (!DWC3_VER_IS_PRIOR(DWC3, 230A)) + reg |=DWC3_DEVTEN_EOPFEN ;
look at cpu usage for dwc3's interrupt before and after this IRQ. Specially when connected to a host that fully supports LPM. IIRC, recent xhci should trigger state changes fairly often.
Still, why do you think you need these events?
Signature · 1 line
balbi
Hi Felipe,
Quoted 10 lines · Felipe Balbi
Quoted 3 lines · Jack Pham
commit
72704f876f50("dwc3: gadget: Implement the suspend entry event handler") introduced (nearly 5 years ago!) an interrupt handler for U3/L1-L2 suspend events. The problem is that these events aren'tlook deeper. They *were* enabled. We've removed because, as it turns out, they just add a TON of interrupts and don't give us much extra information. The only reason why the state change interrupts are still there is because of a known silicon bug in versions < 2.50a. That's all documented in the driver itself.
I did go through the commit history. Are you referring to your change 799e9dc82968 ("usb: dwc3: gadget: conditionally disable Link State change events")? If so then it sounds like you are talking about the link state change event, defined as event value 3 and enabled with DEVTEN bit 3.
The "link state change event" is *not* the same as the one I'm referring to in this patch which is documented in newer revisions of the databook (both DWC3 and DWC3.1) as "USB Suspend Entry" (event 6). It's described as only getting generated when the link enters U3, L2 or L1 states.
Quoted 1 line
For anything that works, we *don't* want link state change interrupts.
Fully agree. I've seen the "link state change" interrupts in action and they are very noisy as they get generated for every single link event, particularly the USB 3.x LTSSM, including U0, U1, U2, Compliance, Polling, etc. in addition to the above U3, L2 & L1.
But I am not proposing re-enabling that particular event type here.
Quoted 6 lines
Quoted 4 lines
currently enabled in the DEVTEN register so the handler is never even invoked. Fix this simply by enabling the corresponding bit in
dwc3_gadget_enable_irq()using the same revision check as found in the handler.More importantly, *why* do you think you need these interrupts?
Bus suspend and resume are useful conditions to be notified about-- that's why we have the .suspend() & .resume() callbacks in struct usb_gadget_driver. But currently the dwc3 gadget does not have any interrupt generated for suspend, and as of now the dwc3_gadget_suspend() does not get called, so it will never invoke the gadget driver's (let's say composite.c) .suspend() routine.
dwc3_gadget_suspend() is called from two places:
1. dwc3_gadget_linksts_change_interrupt() - which is the handler for DWC3_DEVICE_EVENT_LINK_STATUS_CHANGE, the one I believe you are referring to and is only enabled on revisions < 2.50a.
2. dwc3_gadget_suspend_interrupt() - which is the handler for the DWC3_DEVICE_EVENT_EOPF (which I'm promptly renaming to DWC3_DEVICE_EVENT_SUSPEND in patch 2/2)
Quoted 14 lines
Quoted 14 lines
Fixes 72704f876f50 ("dwc3: gadget: Implement the suspend entry event handler")Signed-off-by Jack Pham <[email protected]>--- drivers/usb/dwc3/gadget.c | 4 ++++ 1 file changed, 4 insertions(+)drivers/usb/dwc3/gadget.c +0
-0@@ -2323,6 +2323,10 @@ static void dwc3_gadget_enable_irq(struct dwc3 *dwc) if (DWC3_VER_IS_PRIOR(DWC3, 250A)) reg |=DWC3_DEVTEN_ULSTCNGEN ;
Quoted 7 lines
drivers/usb/dwc3/gadget.c static void dwc3_gadget_enable_irq(struct dwc3 *dwc) · line 2328
Quoted 3 lines
drivers/usb/dwc3/gadget.c +3
-0+/* On 2.30a and above this bit enables U3/L2-L1 Suspend Events */ + if (!DWC3_VER_IS_PRIOR(DWC3, 230A)) + reg |=DWC3_DEVTEN_EOPFEN ;look at cpu usage for dwc3's interrupt before and after this IRQ. Specially when connected to a host that fully supports LPM. IIRC, recent xhci should trigger state changes fairly often.
In our experience for U3 & L2 they are only as frequent as the host enters suspend, which is not that frequent unless the user is toggling system suspend/resume like crazy or has autosuspend enabled with a very short timeout.
You are right that LPM L1 might happen more often when connected to a host that supports it. I'll try to collect some stats in that case to see what is a typical count for a given connection. It's worth noting that the dwc3_gadget_suspend_interrupt() handler specifically checks whether it is U3 (L2 if 2.0) before calling dwc3_gadget_suspend() otherwise it just stores the current link_state and exits. So almost a no-op but not quite since it does incur the cost of an interrupt every time it happens.
Quoted 1 line
Still, why do you think you need these events?
Answered above. I'll add that specifically composite_suspend() also invokes each function driver's .suspend() callback which is needed in case functions have to stop any outstanding endpoint requests and restart them in the .resume() handler. And finally composite_suspend() also invokes the vbus_draw() callback which is needed to properly inform a power supply or charger to stop drawing current to comply with USB suspend power requirements.
Thanks,
Jack
Signature · 2 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project
Hi,
Quoted 20 lines · Jack Pham
Hi Felipe,
Quoted 10 lines · Felipe Balbi
Quoted 3 lines · Jack Pham
commit
72704f876f50("dwc3: gadget: Implement the suspend entry event handler") introduced (nearly 5 years ago!) an interrupt handler for U3/L1-L2 suspend events. The problem is that these events aren'tlook deeper. They *were* enabled. We've removed because, as it turns out, they just add a TON of interrupts and don't give us much extra information. The only reason why the state change interrupts are still there is because of a known silicon bug in versions < 2.50a. That's all documented in the driver itself.
I did go through the commit history. Are you referring to your change
799e9dc82968("usb: dwc3: gadget: conditionally disable Link State change events")? If so then it sounds like you are talking about the link state change event, defined as event value 3 and enabled with DEVTEN bit 3.
Quoted 4 lines
The "link state change event" is *not* the same as the one I'm referring to in this patch which is documented in newer revisions of the databook (both DWC3 and DWC3.1) as "USB Suspend Entry" (event 6). It's described as only getting generated when the link enters U3, L2 or L1 states.
heh, I need some sleep, apparently :-)
Quoted 7 lines
Quoted 6 lines
Quoted 4 lines
currently enabled in the DEVTEN register so the handler is never even invoked. Fix this simply by enabling the corresponding bit in
dwc3_gadget_enable_irq()using the same revision check as found in the handler.More importantly, *why* do you think you need these interrupts?
Quoted 8 lines
Bus suspend and resume are useful conditions to be notified about-- that's why we have the .suspend() & .resume() callbacks in struct
usb_gadget_driver. But currently the dwc3 gadget does not have any interrupt generated for suspend, and as of now thedwc3_gadget_suspend()does not get called, so it will never invoke the gadget driver's (let's saycomposite.c) .suspend() routine.
dwc3_gadget_suspend()is called from two places:
understood.
Quoted 7 lines
1.
dwc3_gadget_linksts_change_interrupt()- which is the handler forDWC3_DEVICE_EVENT_LINK_STATUS_CHANGE, the one I believe you are referring to and is only enabled on revisions < 2.50a.2.
dwc3_gadget_suspend_interrupt()- which is the handler for theDWC3_DEVICE_EVENT_EOPF(which I'm promptly renaming toDWC3_DEVICE_EVENT_SUSPENDin patch 2/2)
Right, now it all makes sense. Thanks for clarifying.
Signature · 1 line
balbi
Quoted 10 lines · Jack Pham
commit
72704f876f50("dwc3: gadget: Implement the suspend entry event handler") introduced (nearly 5 years ago!) an interrupt handler for U3/L1-L2 suspend events. The problem is that these events aren't currently enabled in the DEVTEN register so the handler is never even invoked. Fix this simply by enabling the corresponding bit indwc3_gadget_enable_irq()using the same revision check as found in the handler.Fixes 72704f876f50 ("dwc3: gadget: Implement the suspend entry event handler")Signed-off-by Jack Pham <[email protected]>
Signature · 1 line
balbi