search for: 16core

...ead-after-release by using xadd; > > the prev head value will include the slowpath flag and indicate if we > > need to do PV kicking of suspended spinners -- on modern chips xadd > > isn't (much) more expensive than an add + load. > > > > Result: > > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > > benchmark overcommit %improve > > kernbench 1x -0.13 > > kernbench 2x 0.02 > > dbench 1x -1.77 > > dbench 2x -0.63 > > > > [Jeremy: hinted missing TICKET...

...ead-after-release by using xadd; > > the prev head value will include the slowpath flag and indicate if we > > need to do PV kicking of suspended spinners -- on modern chips xadd > > isn't (much) more expensive than an add + load. > > > > Result: > > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > > benchmark overcommit %improve > > kernbench 1x -0.13 > > kernbench 2x 0.02 > > dbench 1x -1.77 > > dbench 2x -0.63 > > > > [Jeremy: hinted missing TICKET...

...et next to my desk, all four boards. Yes, you could argue that it's merely a toy.... but it's bigger and cheaper than the AWS box that currently hosts my web site, RStudio server, and Shiny server! > Got it. Missed the cluster part earlier and then confused myself looking for arm64 16core machines. There aren't any :) >> For the truly curious, you can view my RPi cluster here: https://photos.app.goo.gl/kALwoYCVxJ32VgxEA > Neat :)? Very geek chic! Patrice [[alternative HTML version deleted]]

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...

...file cabinet next to my desk, all four boards. Yes, you could argue that it's merely a toy.... but it's bigger and cheaper than the AWS box that currently hosts my web site, RStudio server, and Shiny server! Got it. Missed the cluster part earlier and then confused myself looking for arm64 16core machines. There aren't any :) | For the truly curious, you can view my RPi cluster here:?https://photos.app.goo.gl/kALwoYCVxJ32VgxEA Neat :) Very geek chic! | Thanks again, Dirk. Your contributions to Debian and Ubuntu are foundational for my computing platforms. You're too kind. C...

...further avoid the need for a read-after-release by using xadd; > the prev head value will include the slowpath flag and indicate if we > need to do PV kicking of suspended spinners -- on modern chips xadd > isn't (much) more expensive than an add + load. > > Result: > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > benchmark overcommit %improve > kernbench 1x -0.13 > kernbench 2x 0.02 > dbench 1x -1.77 > dbench 2x -0.63 > > [Jeremy: hinted missing TICKET_LOCK_INC for kick] > [Oleg:...

...further avoid the need for a read-after-release by using xadd; > the prev head value will include the slowpath flag and indicate if we > need to do PV kicking of suspended spinners -- on modern chips xadd > isn't (much) more expensive than an add + load. > > Result: > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > benchmark overcommit %improve > kernbench 1x -0.13 > kernbench 2x 0.02 > dbench 1x -1.77 > dbench 2x -0.63 > > [Jeremy: hinted missing TICKET_LOCK_INC for kick] > [Oleg:...

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...

...further avoid the need for a read-after-release by using xadd; > the prev head value will include the slowpath flag and indicate if we > need to do PV kicking of suspended spinners -- on modern chips xadd > isn't (much) more expensive than an add + load. > > Result: > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > benchmark overcommit %improve > kernbench 1x -0.13 > kernbench 2x 0.02 > dbench 1x -1.77 > dbench 2x -0.63 > > [Jeremy: hinted missing TICKET_LOCK_INC for kick] > [Oleg: Mov...

...further avoid the need for a read-after-release by using xadd; > the prev head value will include the slowpath flag and indicate if we > need to do PV kicking of suspended spinners -- on modern chips xadd > isn't (much) more expensive than an add + load. > > Result: > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > benchmark overcommit %improve > kernbench 1x -0.13 > kernbench 2x 0.02 > dbench 1x -1.77 > dbench 2x -0.63 > > [Jeremy: hinted missing TICKET_LOCK_INC for kick] > [Oleg: Mov...

Good morning, in order to post proper documentation, what logs (or log level) do I need to troubleshoot a Centos 7.6.1810 3.10.0-957.27.2.el7.x86_64 tha tis running a NFS server on top LVM on top of XFS on top of VDO on top of MDAM on a 6 SSD disk RAID6 ? This physical NFS server is servign 2 NFS v4.2 shares to 2 physical KVM virtualization hosts. When remote NFS clients start doing intensive

...d; > > > the prev head value will include the slowpath flag and indicate if we > > > need to do PV kicking of suspended spinners -- on modern chips xadd > > > isn't (much) more expensive than an add + load. > > > > > > Result: > > > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > > > benchmark overcommit %improve > > > kernbench 1x -0.13 > > > kernbench 2x 0.02 > > > dbench 1x -1.77 > > > dbench 2x -0.63 > > > > >...

...d; > > > the prev head value will include the slowpath flag and indicate if we > > > need to do PV kicking of suspended spinners -- on modern chips xadd > > > isn't (much) more expensive than an add + load. > > > > > > Result: > > > setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) > > > benchmark overcommit %improve > > > kernbench 1x -0.13 > > > kernbench 2x 0.02 > > > dbench 1x -1.77 > > > dbench 2x -0.63 > > > > >...

...by using xadd; >>> the prev head value will include the slowpath flag and indicate if we >>> need to do PV kicking of suspended spinners -- on modern chips xadd >>> isn't (much) more expensive than an add + load. >>> >>> Result: >>> setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) >>> benchmark overcommit %improve >>> kernbench 1x -0.13 >>> kernbench 2x 0.02 >>> dbench 1x -1.77 >>> dbench 2x -0.63 >>> >>> [Jere...

...by using xadd; >>> the prev head value will include the slowpath flag and indicate if we >>> need to do PV kicking of suspended spinners -- on modern chips xadd >>> isn't (much) more expensive than an add + load. >>> >>> Result: >>> setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) >>> benchmark overcommit %improve >>> kernbench 1x -0.13 >>> kernbench 2x 0.02 >>> dbench 1x -1.77 >>> dbench 2x -0.63 >>> >>> [Jere...

...d for unlock_kick (Linus): We further avoid the need for a read-after-release by using xadd; the prev head value will include the slowpath flag and indicate if we need to do PV kicking of suspended spinners -- on modern chips xadd isn't (much) more expensive than an add + load. Result: setup: 16core (32 cpu +ht sandy bridge 8GB 16vcpu guest) benchmark overcommit %improve kernbench 1x -0.13 kernbench 2x 0.02 dbench 1x -1.77 dbench 2x -0.63 [Jeremy: hinted missing TICKET_LOCK_INC for kick] [Oleg: Moving slowpath flag to head, ticket_equals...