| Message ID | 20210115161907.2875631-2-mathias.nyman@linux.intel.com (mailing list archive) |
|---|---|
| State | Superseded |
| Commit | 576667bad341516edc4e18eb85acb0a2b4c9c9d9 |
| Headers | show |
| Series | xhci fixes for usb-linus | expand |
On 1/15/21 7:19 PM, Mathias Nyman wrote: > Once the command ring doorbell is rung the xHC controller will parse all > command TRBs on the command ring that have the cycle bit set properly. > > If the driver just started writing the next command TRB to the ring when > hardware finished the previous TRB, then HW might fetch an incomplete TRB > as long as its cycle bit set correctly. > > A command TRB is 16 bytes (128 bits) long. > Driver writes the command TRB in four 32 bit chunks, with the chunk > containing the cycle bit last. This does however not guarantee that > chunks actually get written in that order. > > This was detected in stress testing when canceling URBs with several > connected USB devices. > Two consecutive "Set TR Dequeue pointer" commands got queued right > after each other, and the second one was only partially written when > the controller parsed it, causing the dequeue pointer to be set > to bogus values. This was seen as error messages: > > "Mismatch between completed Set TR Deq Ptr command & xHCI internal state" > > Solution is to add a write memory barrier before writing the cycle bit. > > Cc: <stable@vger.kernel.org> > Tested-by: Ross Zwisler <zwisler@google.com> > Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com> > --- > drivers/usb/host/xhci-ring.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c > index 5677b81c0915..cf0c93a90200 100644 > --- a/drivers/usb/host/xhci-ring.c > +++ b/drivers/usb/host/xhci-ring.c > @@ -2931,6 +2931,8 @@ static void queue_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, > trb->field[0] = cpu_to_le32(field1); > trb->field[1] = cpu_to_le32(field2); > trb->field[2] = cpu_to_le32(field3); > + /* make sure TRB is fully written before giving it to the controller */ > + wmb(); Have you tried the lighter barrier, dma_wmb()? IIRC, it exists for these exact cases... > trb->field[3] = cpu_to_le32(field4); > > trace_xhci_queue_trb(ring, trb); MBR, Sergei
From: Sergei Shtylyov > Sent: 15 January 2021 16:40 > > On 1/15/21 7:19 PM, Mathias Nyman wrote: > > > Once the command ring doorbell is rung the xHC controller will parse all > > command TRBs on the command ring that have the cycle bit set properly. > > > > If the driver just started writing the next command TRB to the ring when > > hardware finished the previous TRB, then HW might fetch an incomplete TRB > > as long as its cycle bit set correctly. > > > > A command TRB is 16 bytes (128 bits) long. > > Driver writes the command TRB in four 32 bit chunks, with the chunk > > containing the cycle bit last. This does however not guarantee that > > chunks actually get written in that order. > > > > This was detected in stress testing when canceling URBs with several > > connected USB devices. > > Two consecutive "Set TR Dequeue pointer" commands got queued right > > after each other, and the second one was only partially written when > > the controller parsed it, causing the dequeue pointer to be set > > to bogus values. This was seen as error messages: > > > > "Mismatch between completed Set TR Deq Ptr command & xHCI internal state" > > > > Solution is to add a write memory barrier before writing the cycle bit. > > > > Cc: <stable@vger.kernel.org> > > Tested-by: Ross Zwisler <zwisler@google.com> > > Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com> > > --- > > drivers/usb/host/xhci-ring.c | 2 ++ > > 1 file changed, 2 insertions(+) > > > > diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c > > index 5677b81c0915..cf0c93a90200 100644 > > --- a/drivers/usb/host/xhci-ring.c > > +++ b/drivers/usb/host/xhci-ring.c > > @@ -2931,6 +2931,8 @@ static void queue_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, > > trb->field[0] = cpu_to_le32(field1); > > trb->field[1] = cpu_to_le32(field2); > > trb->field[2] = cpu_to_le32(field3); > > + /* make sure TRB is fully written before giving it to the controller */ > > + wmb(); > > Have you tried the lighter barrier, dma_wmb()? IIRC, it exists for these exact cases... Isn't dma_wmb() needed between the last memory write and the io_write to the doorbell? Here we need to ensure the two memory writes aren't re-ordered. Apart from alpha isn't a barrier() likely to be enough for that. It is worth checking that the failing compiles didn't have the writes reordered. David > > > trb->field[3] = cpu_to_le32(field4); > > > > trace_xhci_queue_trb(ring, trb); > > MBR, Sergei - Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
On 1/15/21 7:50 PM, David Laight wrote: > From: Sergei Shtylyov >> Sent: 15 January 2021 16:40 >> >> On 1/15/21 7:19 PM, Mathias Nyman wrote: >> >>> Once the command ring doorbell is rung the xHC controller will parse all >>> command TRBs on the command ring that have the cycle bit set properly. >>> >>> If the driver just started writing the next command TRB to the ring when >>> hardware finished the previous TRB, then HW might fetch an incomplete TRB >>> as long as its cycle bit set correctly. >>> >>> A command TRB is 16 bytes (128 bits) long. >>> Driver writes the command TRB in four 32 bit chunks, with the chunk >>> containing the cycle bit last. This does however not guarantee that >>> chunks actually get written in that order. >>> >>> This was detected in stress testing when canceling URBs with several >>> connected USB devices. >>> Two consecutive "Set TR Dequeue pointer" commands got queued right >>> after each other, and the second one was only partially written when >>> the controller parsed it, causing the dequeue pointer to be set >>> to bogus values. This was seen as error messages: >>> >>> "Mismatch between completed Set TR Deq Ptr command & xHCI internal state" >>> >>> Solution is to add a write memory barrier before writing the cycle bit. >>> >>> Cc: <stable@vger.kernel.org> >>> Tested-by: Ross Zwisler <zwisler@google.com> >>> Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com> >>> --- >>> drivers/usb/host/xhci-ring.c | 2 ++ >>> 1 file changed, 2 insertions(+) >>> >>> diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c >>> index 5677b81c0915..cf0c93a90200 100644 >>> --- a/drivers/usb/host/xhci-ring.c >>> +++ b/drivers/usb/host/xhci-ring.c >>> @@ -2931,6 +2931,8 @@ static void queue_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, >>> trb->field[0] = cpu_to_le32(field1); >>> trb->field[1] = cpu_to_le32(field2); >>> trb->field[2] = cpu_to_le32(field3); >>> + /* make sure TRB is fully written before giving it to the controller */ >>> + wmb(); >> >> Have you tried the lighter barrier, dma_wmb()? IIRC, it exists for these exact cases... > > Isn't dma_wmb() needed between the last memory write and the io_write to the doorbell? No. > Here we need to ensure the two memory writes aren't re-ordered. No, we need all 3 ring memory writes to be ordered such that they all happen before the 4th write. It's not wonder this bug hasn't been noticed before -- x86 has strong write ordering unlike ARM/etc. > Apart from alpha isn't a barrier() likely to be enough for that. Not sure -- we don't have any barriers before the equivalents of a doorbell write in e.g. the Renesas Ehter driver. > It is worth checking that the failing compiles didn't have the writes reordered. The writes are reordered not because of the compiler -- the read/write reordering is a CPU feature (on at least non-x86). :-) > David > >> >>> trb->field[3] = cpu_to_le32(field4); >>> >>> trace_xhci_queue_trb(ring, trb); MBR, Sergei
On 15.1.2021 19.21, Sergei Shtylyov wrote: > On 1/15/21 7:50 PM, David Laight wrote: >> From: Sergei Shtylyov >>> Sent: 15 January 2021 16:40 >>> >>> On 1/15/21 7:19 PM, Mathias Nyman wrote: >>> >>>> Once the command ring doorbell is rung the xHC controller will parse all >>>> command TRBs on the command ring that have the cycle bit set properly. >>>> >>>> If the driver just started writing the next command TRB to the ring when >>>> hardware finished the previous TRB, then HW might fetch an incomplete TRB >>>> as long as its cycle bit set correctly. >>>> >>>> A command TRB is 16 bytes (128 bits) long. >>>> Driver writes the command TRB in four 32 bit chunks, with the chunk >>>> containing the cycle bit last. This does however not guarantee that >>>> chunks actually get written in that order. >>>> >>>> This was detected in stress testing when canceling URBs with several >>>> connected USB devices. >>>> Two consecutive "Set TR Dequeue pointer" commands got queued right >>>> after each other, and the second one was only partially written when >>>> the controller parsed it, causing the dequeue pointer to be set >>>> to bogus values. This was seen as error messages: >>>> >>>> "Mismatch between completed Set TR Deq Ptr command & xHCI internal state" >>>> >>>> Solution is to add a write memory barrier before writing the cycle bit. >>>> >>>> Cc: <stable@vger.kernel.org> >>>> Tested-by: Ross Zwisler <zwisler@google.com> >>>> Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com> >>>> --- >>>> drivers/usb/host/xhci-ring.c | 2 ++ >>>> 1 file changed, 2 insertions(+) >>>> >>>> diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c >>>> index 5677b81c0915..cf0c93a90200 100644 >>>> --- a/drivers/usb/host/xhci-ring.c >>>> +++ b/drivers/usb/host/xhci-ring.c >>>> @@ -2931,6 +2931,8 @@ static void queue_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, >>>> trb->field[0] = cpu_to_le32(field1); >>>> trb->field[1] = cpu_to_le32(field2); >>>> trb->field[2] = cpu_to_le32(field3); >>>> + /* make sure TRB is fully written before giving it to the controller */ >>>> + wmb(); >>> >>> Have you tried the lighter barrier, dma_wmb()? IIRC, it exists for these exact cases... >> True, good point, dma_wmb() should be enough here. In fact most other wmb()s in xhci could be turned into dma_wmb(). Looks like Greg already picked this so maybe a later patch to usb-next that does this wmb() -> dma_wmb() optimization where possible. >> Isn't dma_wmb() needed between the last memory write and the io_write to the doorbell? > > No. Transfer trbs already have a wmb in giveback_first_trb() So no need in that case. For command trbs it's unlikely but not impossible. The issue we are solving here is xHC controller parsing two commands after a doorbell ring. First one was the intended, properly written command. Second was a out-of order partially written command. driver didn't even ring the doorbell for the second command yet. There are a couple operations between trb last memory write and command doorbell ring. a wmb() in that place would solve a case where memory write is so out of order and delayed that xHC controller reads and reacts to the doorbell ring, and reads the command ring before the memory write to the command ring is done. Unlikely but not impossible. No such issues seen so far, but maybe a dma_wmb() in xhci_ring_cmd_db() wouldn't hurt. -Mathias
On Mon, Jan 18, 2021 at 5:05 AM Mathias Nyman <mathias.nyman@linux.intel.com> wrote: <> > True, good point, dma_wmb() should be enough here. > In fact most other wmb()s in xhci could be turned into dma_wmb(). FWIW I've confirmed in my testing that dma_wmb() does indeed also solve the problem.
diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 5677b81c0915..cf0c93a90200 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -2931,6 +2931,8 @@ static void queue_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, trb->field[0] = cpu_to_le32(field1); trb->field[1] = cpu_to_le32(field2); trb->field[2] = cpu_to_le32(field3); + /* make sure TRB is fully written before giving it to the controller */ + wmb(); trb->field[3] = cpu_to_le32(field4); trace_xhci_queue_trb(ring, trb);