From: "Raj, Ashok" <ashok.raj@intel.com>
To: Stanislav Spassov <stanspas@amazon.com>
Cc: linux-pci@vger.kernel.org,
"Stanislav Spassov" <stanspas@amazon.de>,
linux-acpi@vger.kernel.org, "Bjorn Helgaas" <bhelgaas@google.com>,
"Thomas Gleixner" <tglx@linutronix.de>,
"Andrew Morton" <akpm@linux-foundation.org>,
"Jan H . Schönherr" <jschoenh@amazon.de>,
"Jonathan Corbet" <corbet@lwn.net>,
"Alex Williamson" <alex.williamson@redhat.com>,
"Sinan Kaya" <okaya@kernel.org>,
"Rajat Jain" <rajatja@google.com>,
"kbuild test robot" <lkp@intel.com>,
"Ashok Raj" <ashok.raj@intel.com>
Subject: Re: [PATCH v3 16/17] PCI: Add CRS handling to pci_dev_wait()
Date: Thu, 5 Mar 2020 09:56:20 -0800 [thread overview]
Message-ID: <20200305175620.GA94051@otc-nc-03> (raw)
In-Reply-To: <20200303132852.13184-17-stanspas@amazon.com>
Hi Stanislav,
Thanks for doing this!
On Tue, Mar 03, 2020 at 02:28:51PM +0100, Stanislav Spassov wrote:
> From: Stanislav Spassov <stanspas@amazon.de>
>
> The PCI Express specification dictates minimal amounts of time that the
> host needs to wait after triggering different kinds of resets before it
> is allowed to attempt accessing the device. After this waiting period,
> devices are required to be responsive to Configuration Space reads.
> However, if a device needs more time to actually complete the reset
> operation internally, it may respond to the read with a Completion
> Request Retry Status (CRS), and keep doing so on subsequent reads
> for as long as necessary. If the device is broken, it may even keep
> responding with CRS indefinitely.
>
> The specification also mandates that any Root Port that supports CRS
> and has CRS Software Visibility (CRS SV) enabled will synthesize the
> special value 0x0001 for the Vendor ID and set any other bits to 1
> upon receiving a CRS Completion for a Configuration Read Request that
> includes both bytes of the Vendor ID (offset 0).
>
> IF CRS is supported by Root Port but CRS SV is not enabled, the request
> is retried autonomously by the Root Port. Platform-specific configuration
> registers may exist to limit the number of or time taken by such retries.
Also when CRSV is enabled, config read that doesn't cover VENDOR is
also retried automatically.
>
> If CRS is not supported, or a different register (not Vendor ID) is
> polled, or the device is responding with CA/UR Completions (rather than
> CRS), the behavior is platform-dependent, but generally the Root Port
> synthesizes ~0 to complete the software read.
>
> Previously, pci_dev_wait() avoided taking advantage of CRS. However,
> on platforms where no limit/timeout can be configured as explained
> above, a device responding with CRS for too long (e.g. because it is
> stuck and cannot complete its reset) may trigger more severe error
> conditions (e.g. TOR timeout, 3-strike CPU CATERR), because the Root
> Port never reports back to the lower-level component requesting the
> transaction.
>
> This patch introduces special handling when CRS is available, and
> otherwise falls back to the previous behavior of polling COMMAND.
>
> Signed-off-by: Stanislav Spassov <stanspas@amazon.de>
> ---
> drivers/pci/pci.c | 52 +++++++++++++++++++++++++++++++++++++++--------
> 1 file changed, 44 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
> index f1ba931b0ead..1a504419e0de 100644
> --- a/drivers/pci/pci.c
> +++ b/drivers/pci/pci.c
> @@ -1081,18 +1081,54 @@ static int pci_dev_wait(struct pci_dev *dev, enum pci_init_event event)
> {
> const char *event_name = pci_init_event_name(event);
> int timeout = dev->reset_ready_poll_ms;
> + int waited = 0;
> + int rc = 0;
> +
>
> /*
> * After reset, the device should not silently discard config
> * requests, but it may still indicate that it needs more time by
> - * responding to them with CRS completions. The Root Port will
> - * generally synthesize ~0 data to complete the read (except when
> - * CRS SV is enabled and the read was for the Vendor ID; in that
> - * case it synthesizes 0x0001 data).
> - *
> - * Wait for the device to return a non-CRS completion. Read the
> - * Command register instead of Vendor ID so we don't have to
> - * contend with the CRS SV value.
> + * responding to them with CRS completions. For such completions:
> + * - If CRS SV is enabled on the Root Port, and the read request
> + * covers both bytes of the Vendor ID register, the Root Port
> + * will synthesize the value 0x0001 (and set any extra requested
> + * bytes to 0xff)
> + * - If CRS SV is not enabled on the Root Port, the Root Port must
> + * re-issue the Configuration Request as a new Request.
> + * Depending on platform-specific Root Complex configurations,
> + * the Root Port may stop retrying after a set number of attempts,
> + * or a configured timeout is hit, or continue indefinitely
> + * (ultimately resulting in non-PCI-specific platform errors, such as
> + * a TOR timeout).
> + */
> + if (dev->crssv_enabled) {
> + u32 id;
I like this check to read VENDOR_ID when crssv is enabled. But your patches
seems to define in patch13 and used in patch16?
can we keep them simple? and if possible just this would be a
needed fix. We have some systems that we have found can cause
timeouts if CRSV is enabled, but you read any other register
other than the PCI_VENDOR.
Would prefer to see this fix before the other cleanups can stabilize :-)
I would also mark that for stable.
> +
> + rc = pci_dev_poll_until_not_equal(dev, PCI_VENDOR_ID, 0xffff,
> + 0x0001, event_name, timeout,
> + &waited, &id);
> + if (rc)
> + return rc;
> +
> + /*
> + * If Vendor/Device ID is valid, the device must be ready.
> + * Note: SR-IOV VFs return ~0 for reads to Vendor/Device
> + * ID and will not be recognized as ready by this check.
> + */
> + if (id != 0x0000ffff && id != 0xffff0000 &&
> + id != 0x00000000 && id != 0xffffffff)
> + return 0;
> + }
> +
next prev parent reply other threads:[~2020-03-05 17:56 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-03 13:28 [PATCH v3 00/17] Improve PCI device post-reset readiness polling Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 01/17] PCI: Fall back to slot/bus reset if softer methods timeout Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 02/17] PCI: Remove unused PCI_PM_BUS_WAIT Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 03/17] PCI: Use pci_bridge_wait_for_secondary_bus after SBR Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 04/17] PCI: Do not override delay for D0->D3hot transition Stanislav Spassov
2020-03-03 18:57 ` Rafael J. Wysocki
2020-03-07 10:58 ` Spassov, Stanislav
2020-03-03 13:28 ` [PATCH v3 05/17] PCI: Fix handling of _DSM 8 (avoiding reset delays) Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 06/17] PCI: Fix us->ms conversion in pci_acpi_optimize_delay Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 07/17] PCI: Clean up and document PM/reset delays Stanislav Spassov
2020-03-03 19:03 ` Rafael J. Wysocki
2020-03-07 11:30 ` Spassov, Stanislav
2020-03-03 13:28 ` [PATCH v3 08/17] PCI: Add more delay overrides to struct pci_dev Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 09/17] PCI: Generalize pci_bus_max_d3cold_delay to pci_bus_max_delay Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 10/17] PCI: Use correct delay in pci_bridge_wait_for_secondary_bus Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 11/17] PCI: Refactor pci_dev_wait to remove timeout parameter Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 12/17] PCI: Refactor pci_dev_wait to take pci_init_event Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 13/17] PCI: Cache CRS Software Visibiliy in struct pci_dev Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 14/17] PCI: Introduce per-device reset_ready_poll override Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 15/17] PCI: Refactor polling loop out of pci_dev_wait Stanislav Spassov
2020-03-03 13:28 ` [PATCH v3 16/17] PCI: Add CRS handling to pci_dev_wait() Stanislav Spassov
2020-03-05 17:56 ` Raj, Ashok [this message]
2020-03-06 18:07 ` Spassov, Stanislav
2020-03-03 13:28 ` [PATCH v3 17/17] PCI: Lower PCIE_RESET_READY_POLL_MS from 1m to 1s Stanislav Spassov
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:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200305175620.GA94051@otc-nc-03 \
--to=ashok.raj@intel.com \
--cc=akpm@linux-foundation.org \
--cc=alex.williamson@redhat.com \
--cc=bhelgaas@google.com \
--cc=corbet@lwn.net \
--cc=jschoenh@amazon.de \
--cc=linux-acpi@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=lkp@intel.com \
--cc=okaya@kernel.org \
--cc=rajatja@google.com \
--cc=stanspas@amazon.com \
--cc=stanspas@amazon.de \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).