On 11/22/21 3:02 AM, Christian Brauner wrote:> On Sun, Nov 21, 2021 at 11:17:11AM -0700, Jens Axboe wrote: >> On 11/21/21 10:49 AM, Mike Christie wrote: >>> Convert io_uring and io-wq to use kernel_worker. >> >> I don't like the kernel_worker name, that implies it's always giving you >> a kernel thread or kthread. That's not the io_uring use case, it's >> really just a thread off the original task that just happens to never >> exit to userspace. >> >> Can we do a better name? At least io_thread doesn't imply that. > > Yeah, I had thought about that as well and at first had kernel_uworker() > locally but wasn't convinced. Maybe we should just make it > create_user_worker()?That's better, or maybe even create_user_inkernel_thread() or something? Pretty long, though... I'd be fine with create_user_worker(). -- Jens Axboe
michael.christie at oracle.com
2021-Nov-22 16:47 UTC
[PATCH V5 07/10] io_uring: switch to kernel_worker
On 11/22/21 8:20 AM, Jens Axboe wrote:> On 11/22/21 3:02 AM, Christian Brauner wrote: >> On Sun, Nov 21, 2021 at 11:17:11AM -0700, Jens Axboe wrote: >>> On 11/21/21 10:49 AM, Mike Christie wrote: >>>> Convert io_uring and io-wq to use kernel_worker. >>> >>> I don't like the kernel_worker name, that implies it's always giving you >>> a kernel thread or kthread. That's not the io_uring use case, it's >>> really just a thread off the original task that just happens to never >>> exit to userspace. >>> >>> Can we do a better name? At least io_thread doesn't imply that. >> >> Yeah, I had thought about that as well and at first had kernel_uworker() >> locally but wasn't convinced. Maybe we should just make it >> create_user_worker()? > > That's better, or maybe even create_user_inkernel_thread() or something? > Pretty long, though... I'd be fine with create_user_worker(). >Ok, I'll do: create_user_worker() start_user_worker() since you guys agree. It will also match the PF flag naming. I'll also add more details to the commit message you requested.