PATCH[2.6.32] softirq patch

Hi ,

I work for Montavista (Cavium Inc) as a Technical Lead . I want to push some of the kernel patches to rt community (2.6.32 kernel 2.6.33 rt patch) , so that It will go to the main line These patches are reviewed and approved by our system Architect. I request you to include in the main line . These issues were reported by our customer CISCO.

Problem Description:

x86, OCTEON Plus and Octeon II:

NOHZ: local_softirq_pending 08

Root Cause:
In run_ksoftirqd, _local_bh_enable() is used to enable bottom half. This does not kick off softirq processing.

How Solved:
In run_ksoftirqd, _local_bh_enable() has been replaced with local_bh_enable() to
enable softirq processing .

Here I am attaching the patch for the above bug . If any
questions please contact me at [email protected]
([email protected])

Regards,
Shaiju.

From 2fa7121b912754f8149774d6a012becf542ffa18 Mon Sep 17 00:00:00 2001
From: Sadasivan Shaiju <[email protected]>
Date: Mon, 6 Feb 2012 18:54:53 -0800
Subject: [PATCH] softirq: Call do_softirq() from run_ksoftirqd() when needed

Source: MontaVista Software, LLC
MR: 47371
Type: Defect Fix
Disposition: Local
ChangeID: 3066b1f971e16b3fc22e6c4e4bba9c6786811acc Description:

In run_ksoftirqd _local_bh_enable() has to be replaced with local_bh_enable() to enable softirq processing.

The call to _local_bh_enable() was introduced in the merge of RT patch-2.6.33-rt7.

---
 kernel/softirq.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

index 67116c8..cdf52d2 100644
@@ -1080,7 +1080,7 @@ sleep_more:
 
 			local_irq_disable();
 			per_cpu(softirq_running, cpu) &= ~softirq_mask;
-			_local_bh_enable();
+			local_bh_enable();
 			local_irq_enable();
 
 			cond_resched();

1.7.0.1

Weird, IIRC semiconductor companies switched to Lead Free about 8 years ago.

See Documentation/SubmittingPatches #9

"If the patch does not apply cleanly to the latest kernel version,
Linus will not apply it."

Neither do I. I really give a shit about a Frankenkernel cobbled together by MVista based on 2.6.32 plus a unknown shitload of crap and 2.6.33-whatever-random-patch-version duct-taped on it.

The only way it will go is to /dev/null

Congratulations. Print the approval and pin it to your cubicle wall.

You can request a Pony and you wont get it.

Who cares and how is that relevant to the problem at hand?

Great, so anything else than those three is not affected.

That's a real new information, perhaps not ....

See Documentation/SubmittingPatches #2

See Documentation/SubmittingPatches #2

See Documentation/SubmittingPatches #7

Aside of failing to submit a proper patch:

If this is MVistas answer to the concerns I raised in the last year, then please tell those who tasked you to waste your time on providing completely useless shit to those, who provided the technology on which that useless shit is based on, that they are on the wrong track.

Thanks,

tglx

03.09.2014 02:07, Thomas Gleixner пишет:

Kill'em all! :D

                                                         Pavel.
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Hi Thomas,Pavel,

I will talk to the concerned people in Montavista/cavium and make
sure the relationship improves.

I will make a note of the points you have made.

Regards,
Shaiju.

-----Original Message-----
From: Pavel Vasilyev [mailto:[email protected]]
Sent: Tuesday, September 02, 2014 3:42 PM
To: Thomas Gleixner; Sadasivan Shaiju
Cc: [email protected]; [email protected]
Subject: Re: PATCH[2.6.32] softirq patch

03.09.2014 02:07, Thomas Gleixner пишет:

Kill'em all! :D

                                                          Pavel.
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html