On 08/24/2018 05:54 PM, Waiman Long wrote:
> On 08/23/2018 08:30 PM, Dave Chinner wrote:
>>
>> That's racy. You can't drop the spin lock between
>> xlog_grant_head_wake() and xlog_grant_head_wait(), because
>> free_bytes is only valid while while the spinlock is held. Same for
>> the "wake_all" va
On 08/23/2018 08:30 PM, Dave Chinner wrote:
> On Thu, Aug 23, 2018 at 12:26:10PM -0400, Waiman Long wrote:
>> Running the AIM7 fserver workload on a 2-socket 24-core 48-thread
>> Broadwell system, it was found that there were severe spinlock contention
>> in the XFS code. In particular, native_queu
On Thu, Aug 23, 2018 at 12:26:10PM -0400, Waiman Long wrote:
> Running the AIM7 fserver workload on a 2-socket 24-core 48-thread
> Broadwell system, it was found that there were severe spinlock contention
> in the XFS code. In particular, native_queued_spin_lock_slowpath()
> consumes 69.7% of cpu t
Running the AIM7 fserver workload on a 2-socket 24-core 48-thread
Broadwell system, it was found that there were severe spinlock contention
in the XFS code. In particular, native_queued_spin_lock_slowpath()
consumes 69.7% of cpu time. The xlog_grant_head_check() function call and
its sub-function c
4 matches
Mail list logo
