[dpdk-dev] doc: add known igb_uio issue for i40e

Message ID 1490606177-38274-1-git-send-email-qiming.yang@intel.com (mailing list archive)
State Superseded, archived
Delegated to: Ferruh Yigit
Headers

Checks

Context Check Description
ci/checkpatch success coding style OK
ci/Intel-compilation success Compilation OK

Commit Message

Qiming Yang March 27, 2017, 9:16 a.m. UTC
  When insmod "igb_uio" with "intr_mode=legacy and test link
status interrupt. Since INTx interrupt is not supported by
X710/XL710/XXV710, it will cause Input/Output error when
reading file descriptor.

Signed-off-by: Qiming Yang <qiming.yang@intel.com>
---
 doc/guides/nics/i40e.rst | 13 +++++++++++++
 1 file changed, 13 insertions(+)
  

Comments

Jingjing Wu March 30, 2017, 12:08 p.m. UTC | #1
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Qiming Yang
> Sent: Monday, March 27, 2017 5:16 PM
> To: dev@dpdk.org
> Cc: Yang, Qiming <qiming.yang@intel.com>
> Subject: [dpdk-dev] [PATCH] doc: add known igb_uio issue for i40e
> 
> When insmod "igb_uio" with "intr_mode=legacy and test link status interrupt.
> Since INTx interrupt is not supported by X710/XL710/XXV710, it will cause
> Input/Output error when reading file descriptor.
> 
> Signed-off-by: Qiming Yang <qiming.yang@intel.com>

Acked-by Jingjing Wu <jingjing.wu@intel.com>
  
Ferruh Yigit April 3, 2017, 10:40 a.m. UTC | #2
On 3/27/2017 10:16 AM, Qiming Yang wrote:
> When insmod "igb_uio" with "intr_mode=legacy and test link
> status interrupt. Since INTx interrupt is not supported by
> X710/XL710/XXV710, it will cause Input/Output error when
> reading file descriptor.
> 
> Signed-off-by: Qiming Yang <qiming.yang@intel.com>
> ---
>  doc/guides/nics/i40e.rst | 13 +++++++++++++

Hi John,

There are three different "Known Issue" sections in documentations:

1) Known Issues document, doc/guides/rel_notes/known_issues.rst
2) Release notes, known issues section,
doc/guides/rel_notes/release_17_05.rst
3) Device specific known issues sections, doc/guides/nics/i40e.rst


This patch updates 3), what is the rule on updating those files?


Thanks,
ferruh
  
John McNamara April 3, 2017, 1:30 p.m. UTC | #3
> -----Original Message-----
> From: Yigit, Ferruh
> Sent: Monday, April 3, 2017 11:40 AM
> To: Yang, Qiming <qiming.yang@intel.com>; dev@dpdk.org; Mcnamara, John
> <john.mcnamara@intel.com>; Thomas Monjalon <thomas.monjalon@6wind.com>
> Subject: Re: [dpdk-dev] [PATCH] doc: add known igb_uio issue for i40e
> 
> On 3/27/2017 10:16 AM, Qiming Yang wrote:
> > When insmod "igb_uio" with "intr_mode=legacy and test link status
> > interrupt. Since INTx interrupt is not supported by X710/XL710/XXV710,
> > it will cause Input/Output error when reading file descriptor.
> >
> > Signed-off-by: Qiming Yang <qiming.yang@intel.com>
> > ---
> >  doc/guides/nics/i40e.rst | 13 +++++++++++++
> 
> Hi John,
> 
> There are three different "Known Issue" sections in documentations:
> 
> 1) Known Issues document, doc/guides/rel_notes/known_issues.rst
> 2) Release notes, known issues section,
> doc/guides/rel_notes/release_17_05.rst
> 3) Device specific known issues sections, doc/guides/nics/i40e.rst
> 
> 
> This patch updates 3), what is the rule on updating those files?
> 

Hi,

I was going to make a similar comment on the patchset.

Having "known" issues in three places is a good way of keeping them unknown.

Ideally the process should be:
    1. Add new "known" issues into the release notes.
    2. Move these to known_issues.rst after each release.

This is a little too manual, so perhaps known issues should be added directly 
to known_issues.rst under a header for the release they were identified in.
The release notes could contain a link to known_issues.rst.

However, this patch adds a third location, the NIC document. I have some
sympathy for this approach since the NIC docs generally have a limitations
section which is similar to known issues.

If we agree that this is a good approach then we should move all of the NIC
specific issues to the Limitations sections of the NIC docs and include a
section at the start of the "Known Issues" doc to remind users to also
check the NIC specific doc.

However, I think this needs to be co-ordinated and done in one go. So, for 
this release, let's not move the i40e known issues and instead do a general
cleanup and refactoring of the "Known Issues" at the very start of the 17.08
cycle.

Other opinions?

John
  

Patch

diff --git a/doc/guides/nics/i40e.rst b/doc/guides/nics/i40e.rst
index 043de34..f4073c2 100644
--- a/doc/guides/nics/i40e.rst
+++ b/doc/guides/nics/i40e.rst
@@ -467,3 +467,16 @@  Incorrect Rx statistics when packet is oversize
 When a packet is over maximum frame size, the packet is dropped.
 However the Rx statistics, when calling `rte_eth_stats_get` incorrectly
 shows it as received.
+
+igb uio legacy mode can not be used in X710/XL710/XXV710
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+When insmod ``igb_uio`` with ``intr_mode=legacy`` and test link status
+interrupt, it will cause Input/Output error when reading file descriptor.
+The root cause is X710/XL710/XXV710 NICs lack support for indicating INTx
+is asserted via the interrupt bit in the PCI status register. Linux delect
+them from INTx support table. The related link `linux kernel commit <https:
+//git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/
+drivers/pci/quirks.c?id=8bcf4525c5d43306c5fd07e132bc8650e3491aec>`_.
+
+NOTE: happening when Linux kernel version >= 4.7