LKML Archive on
help / color / mirror / Atom feed
From: "Jonas Dreßler" <>
To: "David Laight" <David.Laight@ACULAB.COM>,
	"'Pali Rohár'" <>
Cc: Amitkumar Karwar <>,
	Ganapathi Bhat <>,
	Xinming Hu <>,
	Kalle Valo <>,
	"David S. Miller" <>,
	Jakub Kicinski <>,
	Tsuchiya Yuto <>,
	"" <>,
	"" <>,
	"" <>,
	"" <>,
	Maximilian Luz <>,
	Andy Shevchenko <>,
	Bjorn Helgaas <>,
	Heiner Kallweit <>,
	Johannes Berg <>,
	Brian Norris <>,
	"" <>
Subject: Re: [PATCH v2 1/2] mwifiex: Use non-posted PCI write when setting TX ring write pointer
Date: Wed, 6 Oct 2021 18:01:02 +0200	[thread overview]
Message-ID: <> (raw)
In-Reply-To: <>

On 9/30/21 16:27, Jonas Dreßler wrote:
> On 9/22/21 5:54 PM, David Laight wrote:
>> From: Pali Rohár
>>> Sent: 22 September 2021 15:27
>>> On Wednesday 22 September 2021 14:03:25 David Laight wrote:
>>>> From: Jonas Dreßler
>>>>> Sent: 14 September 2021 12:48
>>>>> On the 88W8897 card it's very important the TX ring write pointer is
>>>>> updated correctly to its new value before setting the TX ready
>>>>> interrupt, otherwise the firmware appears to crash (probably because
>>>>> it's trying to DMA-read from the wrong place). The issue is present in
>>>>> the latest firmware version 15.68.19.p21 of the pcie+usb card.
>>>>> Since PCI uses "posted writes" when writing to a register, it's not
>>>>> guaranteed that a write will happen immediately. That means the pointer
>>>>> might be outdated when setting the TX ready interrupt, leading to
>>>>> firmware crashes especially when ASPM L1 and L1 substates are enabled
>>>>> (because of the higher link latency, the write will probably take
>>>>> longer).
>>>>> So fix those firmware crashes by always using a non-posted write for
>>>>> this specific register write. We do that by simply reading back the
>>>>> register after writing it, just as a few other PCI drivers do.
>>>>> This fixes a bug where during rx/tx traffic and with ASPM L1 substates
>>>>> enabled (the enabled substates are platform dependent), the firmware
>>>>> crashes and eventually a command timeout appears in the logs.
>>>> I think you need to change your terminology.
>>>> PCIe does have some non-posted write transactions - but I can't
>>>> remember when they are used.
>>> In PCIe are all memory write requests as posted.
>>> Non-posted writes in PCIe are used only for IO and config requests. But
>>> this is not case for proposed patch change as it access only card's
>>> memory space.
>>> Technically this patch does not use non-posted memory write (as PCIe
>>> does not support / provide it), just adds something like a barrier and
>>> I'm not sure if it is really correct (you already wrote more details
>>> about it, so I will let it be).
>>> I'm not sure what is the correct terminology, I do not know how this
>>> kind of write-followed-by-read "trick" is correctly called.
>> I think it is probably best to say:
>>     "flush the posted write when setting the TX ring write pointer".
>> The write can get posted in any/all of the following places:
>> 1) The cpu store buffer.
>> 2) The PCIe host bridge.
>> 3) Any other PCIe bridges.
>> 4) The PCIe slave logic in the target.
>>     There could be separate buffers for each BAR,
>> 5) The actual target logic for that address block.
>>     The target (probably) will look a bit like an old fashioned cpu
>>     motherboard with the PCIe slave logic as the main bus master.
>> The readback forces all the posted write buffers be flushed.
>> In this case I suspect it is either flushing (5) or the extra
>> delay of the read TLP processing that 'fixes' the problem.
>> Note that depending on the exact code and host cpu the second
>> write may not need to wait for the response to the read TLP.
>> So the write, readback, write TLP may be back to back on the
>> actual PCIe link.
>> Although I don't have access to an actual PCIe monitor we
>> do have the ability to trace 'data' TLP into fpga memory
>> on one of our systems.
>> This is near real-time but they are slightly munged.
>> Watching the TLP can be illuminating!
>>     David
>> -
>> Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
>> Registration No: 1397386 (Wales)
> Thanks for the detailed explanations, it looks like indeed the read-back is not the real fix here, a simple udelay(50) before sending the "TX ready" interrupt also does the trick.
>                  } else {
> +                       udelay(50);
> +
>                          /* Send the TX ready interrupt */
>                          if (mwifiex_write_reg(adapter, PCIE_CPU_INT_EVENT,
>                                                CPU_INTR_DNLD_RDY)) {
> I've tested that for a week now and haven't seen any firmware crashes. Interestingly enough it looks like the delay can also be added after setting the "TX ready" interrupt, just not before updating the TX ring write pointer.
> I have no idea if 50 usecs is a good duration to wait here, from trying different values I found that 10 to 20 usecs is not enough, but who knows, maybe that's platform dependent?

So I spent the last few days going slightly crazy while trying to dig deeper
into this.

My theory was that the udelay() delays some subsequent register write or
other communication with the card that would trigger the crash if executed
too early after writing the TX ring write pointer. So I tried moving the
udelay() around, carefully checking when the crash is gone and when it isn't.

In the end my theory turned out completely wrong, what I found was this:
Pinning down the last place where the udelay() is effective gets us here
right before we bail out of the main process and idle.

I tried adding the udelay() as the first thing we do on the next run of the
while-loop after that break, but with that the crash came back.

So what does this mean, we fix the crash by sleeping before idling? Sounds
a bit counterintuitive to me...

The only thing I can take away from this is that maybe the udelay() keeps
the CPU from entering some powersaving state and with that the PCI bus from
entering ASPM states (considering that the crash can also be fixed by
disabling ASPM L1.2).

  reply	other threads:[~2021-10-06 16:01 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-14 11:48 [PATCH v2 0/2] mwifiex: Work around firmware bugs on 88W8897 chip Jonas Dreßler
2021-09-14 11:48 ` [PATCH v2 1/2] mwifiex: Use non-posted PCI write when setting TX ring write pointer Jonas Dreßler
2021-09-22 11:17   ` Andy Shevchenko
2021-09-22 12:08     ` Jonas Dreßler
2021-09-22 13:22       ` Andy Shevchenko
2021-09-22 14:03   ` David Laight
2021-09-22 14:27     ` Pali Rohár
2021-09-22 15:54       ` David Laight
2021-09-30 14:27         ` Jonas Dreßler
2021-10-06 16:01           ` Jonas Dreßler [this message]
2021-09-14 11:48 ` [PATCH v2 2/2] mwifiex: Try waking the firmware until we get an interrupt Jonas Dreßler
2021-09-22 11:19   ` Andy Shevchenko
2021-09-30 18:04     ` Jonas Dreßler
2021-09-30 20:58       ` Andy Shevchenko
2021-09-30 21:07         ` Jonas Dreßler
2021-09-30 21:16           ` Andy Shevchenko
2021-10-03  9:18   ` Jonas Dreßler
2021-10-04 17:52     ` Brian Norris
2021-09-27 20:30 ` [PATCH v2 0/2] mwifiex: Work around firmware bugs on 88W8897 chip Brian Norris

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \ \ \
    --cc=David.Laight@ACULAB.COM \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
    --subject='Re: [PATCH v2 1/2] mwifiex: Use non-posted PCI write when setting TX ring write pointer' \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).