Re: [Regression] Filesystem I/O is CPU-bound in rc7 and rc8

[James Cloos <cloos@jhcloos.com>]

>>>>> "Mike" == Mike Fedyk <mfedyk@mikefedyk.com> writes:

Mike> On Sat, Feb 13, 2010 at 7:11 PM, James Cloos <cloos@jhcloos.com> wrote:
>> Sometime between rc6 and rc7 all filesystem I/O started using 100% CPU,
>> usually on the order of 60% sys, 40% user.
>> 
>> I've tried this with each of ext4, jfs and btrfs filesystems.  All show
>> the same issue.
>> 

Mike> Are you sure you're not running with any of the debugging options
Mike> enabled?  I see the same, but I have debugging enabled (rawhide
Mike> kernel).

There are some:

CONFIG_HAVE_DMA_API_DEBUG=y
CONFIG_X86_DEBUGCTLMSR=y
CONFIG_CFG80211_DEBUGFS=y
CONFIG_MAC80211_DEBUGFS=y
CONFIG_PNP_DEBUG_MESSAGES=y
CONFIG_FIREWIRE_OHCI_DEBUG=y
CONFIG_USB_SERIAL_DEBUG=m
CONFIG_DEBUG_FS=y
CONFIG_DEBUG_KERNEL=y
CONFIG_SCHED_DEBUG=y
CONFIG_DEBUG_BUGVERBOSE=y
CONFIG_DEBUG_MEMORY_INIT=y
CONFIG_DEBUG_STACKOVERFLOW=y
CONFIG_KEYS_DEBUG_PROC_KEYS=y

I forgot about /proc/sched_debug, but am back at -rc6 right now; I'll
post an example of /proc/sched_debug output during the issue when I'm
next in rc8 or later.

-JimC
-- 
James Cloos <cloos@jhcloos.com>         OpenPGP: 1024D/ED7DAEA6