RE: rte_timer_reset related issues in DPDK 20.05

RE: rte_timer_reset related issues in DPDK 20.05

[Nagma Meraj]

[-- Attachment #1.1: Type: text/plain, Size: 1336 bytes --]

TCS Confidential

Hi,

We are working on Front Haul Library which is using DPDK internally for data accleration.
In that we are facing the following issues:
1.In Front Haul Library, one of the threads xran_timing_source_thread() is using DPDK rte_timer_reset() where in the timer_add() it is printing negative values in the timer_get_prev_enteries().
2. That is,in  prev[lvl]->sl_next[lvl] = -459511634 is taking negative value in a while loop as in the below code snippet and is going in an infinite loop because of which the above thread is getting stuck here and it is failing. Can we get possible scenarios when this type of errors generally occurs?

[cid:2e8dda97-6a2a-474d-ba0a-98e7789a4014]


Thanks & Regards,
Nagma Meraj


TCS Confidential
=====-----=====-----=====
Notice: The information contained in this e-mail
message and/or attachments to it may contain 
confidential or privileged information. If you are 
not the intended recipient, any dissemination, use, 
review, distribution, printing or copying of the 
information contained in this e-mail message 
and/or attachments to it are strictly prohibited. If 
you have received this communication in error, 
please notify us by reply e-mail or telephone and 
immediately and permanently delete the message 
and any attachments. Thank you



[-- Attachment #1.2: Type: text/html, Size: 4459 bytes --]

[-- Attachment #2: image.png --]
[-- Type: image/png, Size: 201400 bytes --]

Re: rte_timer_reset related issues in DPDK 20.05

[Stephen Hemminger]

On Mon, 6 Nov 2023 10:31:20 +0000
Nagma Meraj <nagma.meraj@tcs.com> wrote:

> TCS Confidential
> 
> Hi,
> 
> We are working on Front Haul Library which is using DPDK internally for data accleration.
> In that we are facing the following issues:
> 1.In Front Haul Library, one of the threads xran_timing_source_thread() is using DPDK rte_timer_reset() where in the timer_add() it is printing negative values in the timer_get_prev_enteries().

Do you have threads fighting over timer?
What context is rte_timer_reset() being called in?


> 2. That is,in  prev[lvl]->sl_next[lvl] = -459511634 is taking negative value in a while loop as in the below code snippet and is going in an infinite loop because of which the above thread is getting stuck here and it is failing. Can we get possible scenarios when this type of errors generally occurs?
> 
> [cid:2e8dda97-6a2a-474d-ba0a-98e7789a4014]
> 
> 
> Thanks & Regards,
> Nagma Meraj
> 
> 
> TCS Confidential
> =====-----=====-----=====
> Notice: The information contained in this e-mail
> message and/or attachments to it may contain 
> confidential or privileged information. If you are 
> not the intended recipient, any dissemination, use, 
> review, distribution, printing or copying of the 
> information contained in this e-mail message 
> and/or attachments to it are strictly prohibited. If 
> you have received this communication in error, 
> please notify us by reply e-mail or telephone and 
> immediately and permanently delete the message 
> and any attachments. Thank you

Please don't put this on mailing list mail.
If lawyers are watching, in theory we should ignore your mail.

RE: rte_timer_reset related issues in DPDK 20.05

[Jayasvi Battu]

[-- Attachment #1: Type: text/plain, Size: 3105 bytes --]

TCS Confidential


Hi Stephen,

Thank you for your reply.

Please find my answers inline.

1.Do you have threads fighting over timer?

No , we are running the thread on only one core so there is no deadlock here. But as discussed earlier the thread is getting stuck in an infinite while loop with prev[lvl]->sl_next[lvl]  = -459511634 in timer_get_prev_enteries().





2.What context is rte_timer_reset() being called in?



During Uplink and Downlink data transfer, we are using the thread of fronthaul library, where it executes callbacks for every 1 ms, and for every call back we are trying to reset the timer.





3.The commit id"[cid:2e8dda97-6a2a-474d-ba0a-98e7789a4014]" which you have shared, we are unable to find it in the git. If possible, can you please share the link of the same?



--

Thanks and Regards

Jayasvi Battu / Nagma




TCS Confidential

-----Original Message-----
From: Stephen Hemminger <stephen@networkplumber.org>
Sent: Monday, November 6, 2023 9:57 PM
To: Nagma Meraj <nagma.meraj@tcs.com>
Cc: dev@dpdk.org; Jayasvi Battu <jayasvi.battu@tcs.com>; RAJESH GUTTULA <rajesh.guttula@tcs.com>; Nidhi Tomar <nidhi.tomar1@tcs.com>
Subject: Re: rte_timer_reset related issues in DPDK 20.05



External Email : Be cautious while opening any links or attachments.



On Mon, 6 Nov 2023 10:31:20 +0000

Nagma Meraj <nagma.meraj@tcs.com<mailto:nagma.meraj@tcs.com>> wrote:



> TCS Confidential

>

> Hi,

>

> We are working on Front Haul Library which is using DPDK internally for data accleration.

> In that we are facing the following issues:

> 1.In Front Haul Library, one of the threads xran_timing_source_thread() is using DPDK rte_timer_reset() where in the timer_add() it is printing negative values in the timer_get_prev_enteries().



Do you have threads fighting over timer?

What context is rte_timer_reset() being called in?





> 2. That is,in  prev[lvl]->sl_next[lvl] = -459511634 is taking negative value in a while loop as in the below code snippet and is going in an infinite loop because of which the above thread is getting stuck here and it is failing. Can we get possible scenarios when this type of errors generally occurs?

>

> [cid:2e8dda97-6a2a-474d-ba0a-98e7789a4014]

>

>

> Thanks & Regards,

> Nagma Meraj

>

>

> TCS Confidential

> =====-----=====-----=====

> Notice: The information contained in this e-mail message and/or

> attachments to it may contain confidential or privileged information.

> If you are not the intended recipient, any dissemination, use, review,

> distribution, printing or copying of the information contained in this

> e-mail message and/or attachments to it are strictly prohibited. If

> you have received this communication in error, please notify us by

> reply e-mail or telephone and immediately and permanently delete the

> message and any attachments. Thank you



Please don't put this on mailing list mail.

If lawyers are watching, in theory we should ignore your mail.



[-- Attachment #2: Type: text/html, Size: 7650 bytes --]