Bus scan is too chatty

Bus scan is too chatty

[Stephen Hemminger]

DPDK EAL needs to be less verbose for non-existent buses.
Most users won't have embedded devices, and messages like:

EAL: Scan for (fslmc) bus failed.

will be confusing.

Re: Bus scan is too chatty

[Thomas Monjalon]

10/11/2024 20:17, Stephen Hemminger:
> DPDK EAL needs to be less verbose for non-existent buses.
> Most users won't have embedded devices, and messages like:
> 
> EAL: Scan for (fslmc) bus failed.
> 
> will be confusing.

When merging the recent fslmc patches, I've reduced the verbosity
of some fslmc logs appearing at the cleanup stage.

This log is from EAL because fslmc scan fails.
It should not return an error if there is no bus or no environment variable.

Sachin, Hemant, please could you fix it?

RE: Bus scan is too chatty

[Hemant Agrawal]

> -----Original Message-----
> From: Thomas Monjalon <thomas@monjalon.net>
> Sent: Monday, November 11, 2024 3:02 PM
> To: Stephen Hemminger <stephen@networkplumber.org>; Gagandeep Singh
> <G.Singh@nxp.com>; Sachin Saxena (OSS) <sachin.saxena@oss.nxp.com>;
> Hemant Agrawal <hemant.agrawal@nxp.com>
> Cc: dev@dpdk.org
> Subject: Re: Bus scan is too chatty
> Importance: High
> 
> 10/11/2024 20:17, Stephen Hemminger:
> > DPDK EAL needs to be less verbose for non-existent buses.
> > Most users won't have embedded devices, and messages like:
> >
> > EAL: Scan for (fslmc) bus failed.
> >
> > will be confusing.
> 
> When merging the recent fslmc patches, I've reduced the verbosity of some
> fslmc logs appearing at the cleanup stage.
> 
> This log is from EAL because fslmc scan fails.
> It should not return an error if there is no bus or no environment variable.
> 
> Sachin, Hemant, please could you fix it?
> 
[Hemant] yes, let me have a look and provide a fix.

Re: Bus scan is too chatty

[Patrick Robb]

[-- Attachment #1: Type: text/plain, Size: 1594 bytes --]

Hello,

I'm noting here that the compressdev test failures reported over the
weekend for the Community Lab are related to this, as the DTS compressdev
tests fail if EAL contains "failed." So, these patchseries failures can be
ignored by maintainers.

I see the fix from Hemant was merged by Thomas. It looks like some of the
recent failures are from series applied to next-net, so I'm guessing that
branch need to be rebased on main.

Thanks.

On Mon, Nov 11, 2024 at 4:36 AM Hemant Agrawal <hemant.agrawal@nxp.com>
wrote:

>
> > -----Original Message-----
> > From: Thomas Monjalon <thomas@monjalon.net>
> > Sent: Monday, November 11, 2024 3:02 PM
> > To: Stephen Hemminger <stephen@networkplumber.org>; Gagandeep Singh
> > <G.Singh@nxp.com>; Sachin Saxena (OSS) <sachin.saxena@oss.nxp.com>;
> > Hemant Agrawal <hemant.agrawal@nxp.com>
> > Cc: dev@dpdk.org
> > Subject: Re: Bus scan is too chatty
> > Importance: High
> >
> > 10/11/2024 20:17, Stephen Hemminger:
> > > DPDK EAL needs to be less verbose for non-existent buses.
> > > Most users won't have embedded devices, and messages like:
> > >
> > > EAL: Scan for (fslmc) bus failed.
> > >
> > > will be confusing.
> >
> > When merging the recent fslmc patches, I've reduced the verbosity of some
> > fslmc logs appearing at the cleanup stage.
> >
> > This log is from EAL because fslmc scan fails.
> > It should not return an error if there is no bus or no environment
> variable.
> >
> > Sachin, Hemant, please could you fix it?
> >
> [Hemant] yes, let me have a look and provide a fix.
>

[-- Attachment #2: Type: text/html, Size: 2488 bytes --]

Re: Bus scan is too chatty

[Ferruh Yigit]

On 11/12/2024 5:35 PM, Patrick Robb wrote:
> Hello,
> 
> I'm noting here that the compressdev test failures reported over the
> weekend for the Community Lab are related to this, as the DTS
> compressdev tests fail if EAL contains "failed." So, these patchseries
> failures can be ignored by maintainers.
> 
> I see the fix from Hemant was merged by Thomas. It looks like some of
> the recent failures are from series applied to next-net, so I'm guessing
> that branch need to be rebased on main.
> 

Hi Patrick,

I am not clear with problem, but I rebased next-net on -rc2, will this
fix the issue for you?

> Thanks.
> 
> On Mon, Nov 11, 2024 at 4:36 AM Hemant Agrawal <hemant.agrawal@nxp.com
> <mailto:hemant.agrawal@nxp.com>> wrote:
> 
> 
>     > -----Original Message-----
>     > From: Thomas Monjalon <thomas@monjalon.net
>     <mailto:thomas@monjalon.net>>
>     > Sent: Monday, November 11, 2024 3:02 PM
>     > To: Stephen Hemminger <stephen@networkplumber.org
>     <mailto:stephen@networkplumber.org>>; Gagandeep Singh
>     > <G.Singh@nxp.com <mailto:G.Singh@nxp.com>>; Sachin Saxena (OSS)
>     <sachin.saxena@oss.nxp.com <mailto:sachin.saxena@oss.nxp.com>>;
>     > Hemant Agrawal <hemant.agrawal@nxp.com
>     <mailto:hemant.agrawal@nxp.com>>
>     > Cc: dev@dpdk.org <mailto:dev@dpdk.org>
>     > Subject: Re: Bus scan is too chatty
>     > Importance: High
>     >
>     > 10/11/2024 20:17, Stephen Hemminger:
>     > > DPDK EAL needs to be less verbose for non-existent buses.
>     > > Most users won't have embedded devices, and messages like:
>     > >
>     > > EAL: Scan for (fslmc) bus failed.
>     > >
>     > > will be confusing.
>     >
>     > When merging the recent fslmc patches, I've reduced the verbosity
>     of some
>     > fslmc logs appearing at the cleanup stage.
>     >
>     > This log is from EAL because fslmc scan fails.
>     > It should not return an error if there is no bus or no environment
>     variable.
>     >
>     > Sachin, Hemant, please could you fix it?
>     >
>     [Hemant] yes, let me have a look and provide a fix.
> 

Re: Bus scan is too chatty

[Patrick Robb]

[-- Attachment #1: Type: text/plain, Size: 370 bytes --]

No worries Ferruh, the reason is not very interesting so you can stop here.
:) Thanks for rebasing on rc2. But for the record, the cause is the EAL
message:

"EAL: Scan for (fslmc) bus failed."

Which I believe is now removed with this commit which is now in next-net:
https://git.dpdk.org/next/dpdk-next-net/commit/?id=b472c50aeee1ca4f3a86be0580d31474653dfd8f

Cheers.

[-- Attachment #2: Type: text/html, Size: 619 bytes --]