Project

General

Profile

Actions
Copy link

Bug #1049

open

Incorrect order of irq function calls for irq_chip

Added by Alexey Khoroshilov over 13 years ago. Updated almost 13 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
Ilja Zakharov
Category:
Environment Generation
Start date:
04/07/2011
Due date:
% Done:

0%

Estimated time:
Detected in build:
svn
Platform:
Published in build:

Description

Many false positives come from incorrect order of irq_chip operations:

static struct irq_chip ab8500_irq_chip = { 
 .name = "ab8500",
 .irq_bus_lock = ab8500_irq_lock,
 .irq_bus_sync_unlock = ab8500_irq_sync_unlock,
 .irq_mask = ab8500_irq_mask,
 .irq_unmask = ab8500_irq_unmask,
 };

irq_bus_lock() and irq_bus_sync_unlock() have to be interleaved.

Actions
Copy link
 #1

Updated by Pavel Shved over 13 years ago

  • Category set to Environment Generation
  • Status changed from New to Open
  • Assignee set to Vadim Mutilin

Vadim says that the issue is a bug indeed.

Actions
Copy link
 #2

Updated by Ilja Zakharov about 13 years ago

  • Assignee changed from Vadim Mutilin to Ilja Zakharov

Problem appears for example on drivers/mfd/ab8500-core.c, but for this driver (and others in mfd directory) can't find .ko files. Starting ldv-manager with all directory don't give anything about this.

Actions
Copy link
 #3

Updated by Vadim Mutilin about 13 years ago

I think the bug was detected by ldv-git

Actions
Copy link
 #4

Updated by Ilja Zakharov about 13 years ago

  • Status changed from Open to Resolved

Driver compiled separately from kernel, false-positives were detected, then i added needed pattern to config file of environment generator.So, it looks like problem is resolved.

Actions
Copy link
 #5

Updated by Evgeny Novikov almost 13 years ago

Alexey, is it true?

Actions
Copy link

Also available in: Atom PDF